feat(omp): bridge hook-guard denials and context through native hooks - #3506
andrebrait wants to merge 20 commits into
Conversation
![graphify-labs[bot]](https://avatars.githubusercontent.com/in/4356270?s=60&v=4){kind=small}
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 2 advisory finding(s) below merit a look before merge.
Formal verification. 2 change(s) tested, no difference found (not proven).
Graphify review — findings
Adds a native Oh My Pi guard integration via graphify omp install, which delegates registration to omp plugin install on the shipped graphify/omp package directory (with graphify omp path printing that directory for manual linking); the omp command is silenced from the stale-graph warning and rejects args other than install/path. The graphify TypeScript extension runs the installed hook-guard read|search CLI before native read/glob/grep/search bash calls on local paths, turning policy denials into OMP block/reason and guidance into a single deduplicated context message that resets on new user runs and session boundaries; it resolves graphify only from an absolute PATH entry outside the project, caps input/output/time per call, and fails open on any error. Trust gating relies on ctx.isProjectTrusted(), which current OMP reports as always-true, so it is not an enforced sandbox — OMP links must be removed with omp plugin uninstall graphify-omp since graphify uninstall does not touch OMP's registry.
Worth a look
- Project-controlled PATH symlink can bypass graphify executable rejection —
graphify/omp/index.ts:28· Escalate · high- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Shared abort controller lets one session cancel another session's guard —
graphify/omp/index.ts:56· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 680 functions depend on the 413 functions this change touches.
Health — this change adds coupling hotspots:
- new:
dispatch_command()— 2 callers, 124 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
claude_install()— 19 callers, 4 callees - new:
gemini_install()— 10 callers, 7 callees - new:
claude_uninstall()— 17 callers, 4 callees - new:
_project_uninstall()— 5 callers, 13 callees - new:
_copy_skill_file()— 12 callers, 5 callees - …and 13 more — each is listed as a finding
Verification — 680 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 680 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
270 of 270 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_member_calls.py— full-run-safetytests/test_cross_repo_shared_types.py— full-run-safetytests/test_csharp_call_site_generic_args.py— full-run-safetytests/test_csharp_enum_members.py— full-run-safety- … and 220 more
non-code file(s) changed (
CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
CHANGELOG.md,README.md,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Formal verification
No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).
The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).
The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
· 21 more finding(s) on lines outside this diff (see the check run).
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 5 advisory finding(s) below merit a look before merge.
Formal verification. 2 change(s) tested, no difference found (not proven).
Graphify review — findings
Adds a native Oh My Pi guard integration via graphify omp install, which links the shipped graphify/omp package (containing package.json and a TypeScript index.ts extension) through omp plugin install, with graphify omp path printing the package directory and both commands routed through dispatch_install_cli and silenced from the stale-graph warning. The extension runs the installed graphify hook-guard read|search CLI before native read/glob/grep/search bash calls, turning policy denials into OMP block/reason and guidance into one deduplicated per-run context message; it only inspects local filesystem targets, requires an absolute-PATH graphify outside the project, and fails open on missing commands, oversized I/O, or a two-second timeout. Note the guard relies on ctx.isProjectTrusted(), which current OMP always returns true, so it is not an enforced sandbox and should only be enabled in trusted projects.
Worth a look
- Command inside project cwd is accepted, contradicting stated security intent —
graphify/omp/index.ts:31· Escalate · high- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Read guard payload ignores the actual file_path input —
graphify/omp/index.ts:79· Escalate · high- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Glob pattern is overwritten with resolved directory —
graphify/omp/index.ts:92· Escalate · high- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Installer executes first omp found on PATH —
graphify/install.py:2113· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Concurrent tool_call handlers race on shared guidance state via check-then-act —
graphify/omp/index.ts:84· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 680 functions depend on the 413 functions this change touches.
Health — this change adds coupling hotspots:
- new:
dispatch_command()— 2 callers, 124 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
claude_install()— 19 callers, 4 callees - new:
gemini_install()— 10 callers, 7 callees - new:
claude_uninstall()— 17 callers, 4 callees - new:
_project_uninstall()— 5 callers, 13 callees - new:
_copy_skill_file()— 12 callers, 5 callees - …and 13 more — each is listed as a finding
Verification — 680 functions in the blast radius were not formally verified this run (proofs are advisory here).
Health delta baseline: last indexed commit fe66389 (diverged from this PR's base — delta is approximate).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 680 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
270 of 270 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_member_calls.py— full-run-safetytests/test_cross_repo_shared_types.py— full-run-safetytests/test_csharp_call_site_generic_args.py— full-run-safetytests/test_csharp_enum_members.py— full-run-safety- … and 220 more
non-code file(s) changed (
CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Formal verification
No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).
The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).
The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
· 21 more finding(s) on lines outside this diff (see the check run).
|
Triage of the refreshed automated review on
No code changes for those five. Dogfooding did find one real packaging issue: replaying the PR onto a newer Graphify release can leave |
37dd4ac to
445ee20
Compare
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 2 advisory finding(s) below merit a look before merge.
Formal verification. 2 change(s) tested, no difference found (not proven).
Graphify review — findings
Adds a native Oh My Pi guard integration exposed as graphify omp install, which delegates registration to omp plugin install on the wheel-shipped graphify/omp package directory (graphify omp path prints that directory for manual linking); it refuses unless omp is on PATH and rejects any argument other than install/path. The shipped TypeScript extension runs the installed graphify hook-guard read|search CLI before native read/glob/grep/search-bash calls, turning policy denials into OMP block/reason and guidance into a deduplicated, per-run/session context message, with in-flight hooks cancelled on session boundaries. It resolves graphify only from absolute PATH entries outside the project (no project-Python fallback), inspects local paths only, and fails open under a 256 KiB input / 64 KiB output / 2s-kill budget; omp is also added to the silent-command set so it skips the freshness warning.
Worth a look
- Read tool ignores file_path input —
graphify/omp/index.ts:79· Escalate · high- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Grep break occurs after processing only first non-remote path, but expandDelimitedPathEntries may yield multiple search roots —
graphify/omp/index.ts:111· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 691 functions depend on the 418 functions this change touches.
Health — this change adds coupling hotspots:
- new:
dispatch_command()— 2 callers, 125 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
claude_install()— 19 callers, 4 callees - new:
_copy_skill_file()— 12 callers, 6 callees - new:
gemini_install()— 10 callers, 7 callees - new:
claude_uninstall()— 17 callers, 4 callees - new:
_project_uninstall()— 5 callers, 13 callees - …and 14 more — each is listed as a finding
Verification — 691 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 691 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
288 of 288 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safetytests/test_cross_repo_member_calls.py— full-run-safetytests/test_cross_repo_shared_types.py— full-run-safety- … and 238 more
non-code file(s) changed (
CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Formal verification
No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).
The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).
The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
· 22 more finding(s) on lines outside this diff (see the check run).
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 5 advisory finding(s) below merit a look before merge.
Formal verification. 2 change(s) tested, no difference found (not proven).
Graphify review — findings
Adds a native Oh My Pi guard integration installable via graphify omp install, which shells out to omp plugin install against the shipped graphify/omp package directory (omp path prints that directory for manual linking); it registers the omp subcommand in dispatch_install_cli and marks it silent so the freshness check is skipped. The shipped extension runs the existing graphify hook-guard read|search CLI before native read/glob/grep/search-bash calls, turning policy denials into OMP block/reason and guidance into a deduplicated per-run context message, inspecting only local filesystem targets with 256 KiB input / 64 KiB output caps and a two-second subprocess budget, and failing open on missing commands or bad output. Adds a CI omp-bridge job that runs tests/omp.test.ts under bun against the installed CLI, gitignores graphify/omp/node_modules/, and documents that removal goes through omp plugin uninstall graphify-omp rather than graphify uninstall.
Worth a look
- Read tool ignores file_path and guards cwd instead —
graphify/omp/index.ts:79· Escalate · high- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Read/Glob guard ignores native path fields —
graphify/omp/index.ts:79· Escalate · high- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Stale generation write into shared guidance set after reset —
graphify/omp/index.ts· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Oversized tool input fails open —
graphify/omp/index.ts:75· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- GRAPHIFY_TEST_CLI can recurse after PATH is replaced —
tests/omp.test.ts:64· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 699 functions depend on the 426 functions this change touches.
Health — this change adds coupling hotspots:
- new:
dispatch_command()— 2 callers, 125 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
claude_install()— 19 callers, 4 callees - new:
_copy_skill_file()— 12 callers, 6 callees - new:
gemini_install()— 10 callers, 7 callees - new:
claude_uninstall()— 17 callers, 4 callees - new:
_project_uninstall()— 5 callers, 13 callees - …and 14 more — each is listed as a finding
Verification — 699 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 699 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
288 of 288 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safetytests/test_cross_repo_member_calls.py— full-run-safetytests/test_cross_repo_shared_types.py— full-run-safety- … and 238 more
non-code file(s) changed (
.github/workflows/ci.yml,.gitignore,CHANGELOG.md,README.md,graphify/omp/bun.lock…) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
.github/workflows/ci.yml,CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Formal verification
No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).
The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).
The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
· 22 more finding(s) on lines outside this diff (see the check run).
585baed to
f38f25b
Compare
|
Rebased onto the latest Since the last push, the OMP bridge tests ( The upstream checks for this PR are sitting in |
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 2 advisory finding(s) below merit a look before merge.
Formal verification. 2 change(s) tested, no difference found (not proven).
Graphify review — findings
Adds a native Oh My Pi guard integration behind graphify omp: graphify omp install delegates to omp plugin install on the shipped graphify/omp/ package directory (erroring if omp is absent), and graphify omp path prints that directory for manual linking. The shipped OMP extension runs the installed graphify hook-guard read|search CLI before native read/glob/grep/search-bash calls, translating the existing Python policy's decisions into OMP block/reason and deduplicated per-run context guidance under 256 KiB/64 KiB/2s caps that fail open, and requires graphify on an absolute PATH outside the project with no fallback. Registers omp as a silent install command, wires an omp-bridge CI job that runs tests/omp.test.ts against the pinned @oh-my-pi/pi-coding-agent devDependency and the real installed CLI, and documents that OMP links must be removed via omp plugin uninstall graphify-omp before uninstalling the Python package.
Worth a look
- Generation check bypasses guidance mutation after await —
graphify/omp/index.ts· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Guard bypass: 'Bash' tool payload never has command inspected, only path —
graphify/omp/index.ts· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 699 functions depend on the 426 functions this change touches.
Health — this change adds coupling hotspots:
- new:
dispatch_command()— 2 callers, 125 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
claude_install()— 19 callers, 4 callees - new:
_copy_skill_file()— 12 callers, 6 callees - new:
gemini_install()— 10 callers, 7 callees - new:
claude_uninstall()— 17 callers, 4 callees - new:
_project_uninstall()— 5 callers, 13 callees - …and 14 more — each is listed as a finding
Verification — 699 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 699 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
288 of 288 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safetytests/test_cross_repo_member_calls.py— full-run-safetytests/test_cross_repo_shared_types.py— full-run-safety- … and 238 more
non-code file(s) changed (
.github/workflows/ci.yml,.gitignore,CHANGELOG.md,README.md,graphify/omp/bun.lock…) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
.github/workflows/ci.yml,CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Formal verification
No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).
The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).
The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
· 22 more finding(s) on lines outside this diff (see the check run).
a99e1ac to
50b4d73
Compare
|
Pushed Change went through a condensed adversarial review round before push. Findings addressed: multi-target accumulation (fixed + regression test with a varying-output fixture), and a trust recheck at delivery time (restored the distrust watchdog the context handler used to provide). Declined, with rationale: restoring the cumulative Bun suite: 7 pass / 0 fail (64 assertions) against the real installed CLI. Fork CI: https://github.com/pfBlockerNG/graphify/actions — upstream runs still need a maintainer's workflow approval. |
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 1 advisory finding(s) below merit a look before merge.
Formal verification. 2 change(s) tested, no difference found (not proven).
Graphify review — findings
Adds a native Oh My Pi guard integration via graphify omp install, which delegates to omp plugin install against the shipped graphify/omp package directory; graphify omp path prints that directory for manual linking, and both commands are silenced from the freshness-warning check. Before native read/glob/grep and search-style bash calls the extension runs the installed graphify hook-guard CLI to reuse the existing fresh/stale policy — turning block/reason and per-call guidance appended to tool results, with strict-mode denials gated on GRAPHIFY_HOOK_STRICT=1 — and only inspects local filesystem targets, with size/time caps that fail open on missing commands or errors. Adds a CI omp-bridge job that installs both uv and bun to run tests/omp.test.ts against the real CLI and the pinned @oh-my-pi/pi-coding-agent devDependency, and ignores the extension's node_modules.
Worth a look
- OMP installer executes first
ompfound on PATH —graphify/install.py:2121· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 699 functions depend on the 426 functions this change touches.
Health — this change adds coupling hotspots:
- new:
dispatch_command()— 2 callers, 125 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
claude_install()— 19 callers, 4 callees - new:
_copy_skill_file()— 12 callers, 6 callees - new:
gemini_install()— 10 callers, 7 callees - new:
claude_uninstall()— 17 callers, 4 callees - new:
_project_uninstall()— 5 callers, 13 callees - …and 14 more — each is listed as a finding
Verification — 699 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 699 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
288 of 288 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safetytests/test_cross_repo_member_calls.py— full-run-safetytests/test_cross_repo_shared_types.py— full-run-safety- … and 238 more
non-code file(s) changed (
.github/workflows/ci.yml,.gitignore,CHANGELOG.md,README.md,graphify/omp/bun.lock…) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
.github/workflows/ci.yml,CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Formal verification
No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).
The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).
The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
· 22 more finding(s) on lines outside this diff (see the check run).
|
Pushed a fix (2904921) for a load-bearing import break plus a CI hardening step. The break: Why CI didn't catch it: the The fix:
Verified: |
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 1 advisory finding(s) below merit a look before merge.
Formal verification. 2 change(s) tested, no difference found (not proven).
Graphify review — findings
Adds a native Oh My Pi integration: graphify omp install delegates to omp plugin install on the shipped graphify/omp package directory (erroring if omp isn't on PATH), and graphify omp path prints that directory for manual linking. Before native read/glob/grep/search-bash calls the OMP extension runs the installed graphify hook-guard CLI against local filesystem targets only, turning policy decisions into block/reason denials and appending per-call guidance to tool results, with 256 KiB input / 64 KiB output caps, a two-second subprocess budget, and fail-open on any error — no graph is built or updated. Adds a CI omp-bridge job that runs tests/omp.test.ts against both the pinned OMP devDeps (blocking) and @latest (advisory, continue-on-error) to catch upstream symbol moves.
Worth a look
- PATH-resolved hook executable inherits full secret environment —
graphify/omp/index.ts:28· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 701 functions depend on the 428 functions this change touches.
Health — this change adds coupling hotspots:
- new:
dispatch_command()— 2 callers, 125 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
claude_install()— 19 callers, 4 callees - new:
_copy_skill_file()— 12 callers, 6 callees - new:
gemini_install()— 10 callers, 7 callees - new:
claude_uninstall()— 17 callers, 4 callees - new:
_project_uninstall()— 5 callers, 13 callees - …and 14 more — each is listed as a finding
Verification — 701 functions in the blast radius were not formally verified this run (proofs are advisory here).
Health delta baseline: last indexed commit b9cd957 (diverged from this PR's base — delta is approximate).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 701 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
288 of 288 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safetytests/test_cross_repo_member_calls.py— full-run-safetytests/test_cross_repo_shared_types.py— full-run-safety- … and 238 more
non-code file(s) changed (
.github/workflows/ci.yml,.gitignore,CHANGELOG.md,README.md,graphify/omp/bun.lock…) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
.github/workflows/ci.yml,CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Formal verification
No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).
The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).
The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
· 22 more finding(s) on lines outside this diff (see the check run).
|
Pushed a guard fix + bridge simplification on top of this branch (SHA Guard fix ( Bridge simplification ( Full test coverage (new |
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 5 advisory finding(s) below merit a look before merge.
Formal verification. 4 change(s) tested, no difference found (not proven).
Graphify review — findings
Adds an omp command that installs a native Oh My Pi guard package via OMP's supported omp plugin install <dir> route (with an omp path subcommand to print the shipped directory), and suppresses Graphify's staleness warning for omp runs. The OMP extension runs the existing hook-guard read|search CLI before native read/glob/grep/search-bash calls, turning policy denials into OMP block/reason and appending per-call guidance to tool results, while inspecting only local filesystem targets and failing open on missing commands, bad output, or timeouts. Tightens hook-guard path classification so a bare www.host/... value counts as remote only when no identically named local path exists (mirroring OMP's local-wins precedence), and wires up a CI omp-bridge job that runs the bridge tests against both the pinned and latest OMP versions, the latter advisory-only so an upstream release can't block a merge.
Worth a look
- file:// authority is dropped during hook path normalization —
graphify/cli.py:983· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Hook guard executable is selected from PATH —
graphify/omp/index.ts:38· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Runtime OMP imports are declared only as devDependencies —
graphify/omp/package.json:8· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Read input ignores file_path and falls back to project root —
graphify/omp/index.ts:92· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- CI executes unpinned latest npm packages —
.github/workflows/ci.yml:116· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 833 functions depend on the 520 functions this change touches.
Health — this change adds coupling hotspots:
- new:
dispatch_command()— 2 callers, 125 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
claude_install()— 19 callers, 4 callees - new:
_copy_skill_file()— 12 callers, 6 callees - new:
gemini_install()— 10 callers, 7 callees - new:
claude_uninstall()— 17 callers, 4 callees - new:
_project_uninstall()— 5 callers, 13 callees - …and 16 more — each is listed as a finding
Verification — 833 functions in the blast radius were not formally verified this run (proofs are advisory here).
Health delta baseline: last indexed commit b9cd957 (diverged from this PR's base — delta is approximate).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 833 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
289 of 289 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safetytests/test_cross_repo_member_calls.py— full-run-safetytests/test_cross_repo_shared_types.py— full-run-safety- … and 239 more
non-code file(s) changed (
.github/workflows/ci.yml,.gitignore,CHANGELOG.md,README.md,graphify/omp/bun.lock…) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
.github/workflows/ci.yml,CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Formal verification
No difference found (not proven): No behavior difference found in \_is\_cwd\_relative (not a proof).
The verifier ran both versions of \_is\_cwd\_relative on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_hook\_guard (not a proof).
The verifier ran both versions of \_run\_hook\_guard on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).
The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).
The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
· 24 more finding(s) on lines outside this diff (see the check run).
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 2 advisory finding(s) below merit a look before merge.
Formal verification. 4 change(s) tested, no difference found (not proven).
Graphify review — findings
Adds a native Oh My Pi guard integration: graphify omp install links the shipped graphify/omp package via OMP's omp plugin install, and graphify omp path prints that directory for manual linking. The extension runs the installed graphify hook-guard read|search CLI before native read/glob/grep/search-bash calls, turning strict-mode policy denials into OMP block/reason and appending per-call guidance to each tool result — inspecting only local filesystem targets and failing open on missing commands, oversized I/O, or the two-second budget. Silences the update-nudge for omp commands, teaches the search-scope guard to treat a bare www.host/... target as remote only when no identically-named local path exists (_is_external_www_target), and gates the new bridge behind a dedicated CI job that runs tests/omp.test.ts against both the pinned and latest OMP (the latter advisory).
Worth a look
- Read tool target is ignored when input uses file_path —
graphify/omp/index.ts:104· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- CI executes mutable latest npm package code —
.github/workflows/ci.yml:123· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 832 functions depend on the 519 functions this change touches.
Health — this change adds coupling hotspots:
- new:
dispatch_command()— 2 callers, 125 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
claude_install()— 19 callers, 4 callees - new:
_copy_skill_file()— 12 callers, 6 callees - new:
gemini_install()— 10 callers, 7 callees - new:
claude_uninstall()— 17 callers, 4 callees - new:
_project_uninstall()— 5 callers, 13 callees - …and 16 more — each is listed as a finding
Verification — 832 functions in the blast radius were not formally verified this run (proofs are advisory here).
Health delta baseline: last indexed commit b9cd957 (diverged from this PR's base — delta is approximate).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 832 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
289 of 289 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safetytests/test_cross_repo_member_calls.py— full-run-safetytests/test_cross_repo_shared_types.py— full-run-safety- … and 239 more
non-code file(s) changed (
.github/workflows/ci.yml,.gitignore,CHANGELOG.md,README.md,graphify/omp/bun.lock…) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
.github/workflows/ci.yml,CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Formal verification
No difference found (not proven): No behavior difference found in \_is\_cwd\_relative (not a proof).
The verifier ran both versions of \_is\_cwd\_relative on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_hook\_guard (not a proof).
The verifier ran both versions of \_run\_hook\_guard on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).
The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).
The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
· 24 more finding(s) on lines outside this diff (see the check run).
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 4 advisory finding(s) below merit a look before merge.
Formal verification. 4 change(s) tested, no difference found (not proven).
Graphify review — findings
Adds a native Oh My Pi guard integration exposed via graphify omp install (delegating to omp plugin install on the shipped graphify/omp/ package) and graphify omp path, which runs the installed hook-guard before native read/glob/grep/search-bash calls, turns policy denials into OMP block/reason, and appends per-call guidance to tool results without ever building or updating a graph. Tightens remote-vs-local path handling in _run_hook_guard so a bare www.host/... target is treated as external only when no identically named local path exists (mirroring OMP's resolveToolSearchScope precedence), and vendors isReadableUrlPath locally in graphify/omp/index.ts rather than importing it from an upstream package that has moved the symbol between releases. Wires up CI with an omp-bridge job that runs the bun bridge tests against both the pinned OMP devDependency (blocking) and the latest published OMP (advisory), and silences the startup version check for omp commands. Note: the extension checks ctx.isProjectTrusted(), but current OMP hardcodes that to true, so this is trust-gated in name only — enable only in trusted projects, and graphify uninstall does not remove the OMP link (omp plugin uninstall graphify-omp must run first).
Worth a look
- CI installs mutable npm package code at @latest —
.github/workflows/ci.yml:124· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Remote file:// authorities bypass URL guard —
graphify/omp/index.ts:41· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Read events ignore file_path and inspect cwd instead —
graphify/omp/index.ts:107· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- file:// host component is silently discarded —
graphify/cli.py:985· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 832 functions depend on the 519 functions this change touches.
Health — this change adds coupling hotspots:
- new:
dispatch_command()— 2 callers, 125 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
claude_install()— 19 callers, 4 callees - new:
_copy_skill_file()— 12 callers, 6 callees - new:
gemini_install()— 10 callers, 7 callees - new:
claude_uninstall()— 17 callers, 4 callees - new:
_project_uninstall()— 5 callers, 13 callees - …and 16 more — each is listed as a finding
Verification — 832 functions in the blast radius were not formally verified this run (proofs are advisory here).
Health delta baseline: last indexed commit b9cd957 (diverged from this PR's base — delta is approximate).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 832 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
289 of 289 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safetytests/test_cross_repo_member_calls.py— full-run-safetytests/test_cross_repo_shared_types.py— full-run-safety- … and 239 more
non-code file(s) changed (
.github/workflows/ci.yml,.gitignore,CHANGELOG.md,README.md,graphify/omp/bun.lock…) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
.github/workflows/ci.yml,CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Formal verification
No difference found (not proven): No behavior difference found in \_is\_cwd\_relative (not a proof).
The verifier ran both versions of \_is\_cwd\_relative on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_hook\_guard (not a proof).
The verifier ran both versions of \_run\_hook\_guard on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).
The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).
The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
· 24 more finding(s) on lines outside this diff (see the check run).
|
Thanks for the review — going through all four findings below. Two were genuine bugs (both stemming from the same root cause), one CI hardening item is now addressed differently, and one is refuted with evidence. Appreciate the flag on all of them; the bot was right to mark them unverified and ask for human review. #2 and #4 — remote
|
fd869d8 to
22caf5e
Compare
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 3 advisory finding(s) below merit a look before merge.
Formal verification. 4 change(s) tested, no difference found (not proven).
Graphify review — findings
Adds an Oh My Pi native guard shipped as a linkable package, installed via graphify omp install (with graphify omp path for manual linking and omp plugin uninstall graphify-omp to remove it). Before native read/glob/grep/search bash calls the extension runs the installed graphify hook-guard CLI to reuse the existing fresh/stale policy for bounded denials and per-call guidance appended to tool results, inspecting only local filesystem targets with 256 KiB input / 64 KiB output caps and a two-second subprocess budget, and failing open on missing commands or bad output. Adds an omp-bridge CI job that runs the bridge tests against the pinned OMP plus a non-blocking advisory step that statically checks whether @latest still exports the path-utils symbols index.ts imports.
Worth a look
- OMP guard ignores Read/Glob canonical path fields —
graphify/omp/index.ts:111· Escalate · high- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Import allow-list test misses side-effect imports —
tests/omp.test.ts· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Test leaks mutated PATH into the rest of the test process —
tests/omp.test.ts· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 842 functions depend on the 529 functions this change touches.
Health — this change adds coupling hotspots:
- new:
dispatch_command()— 2 callers, 125 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
claude_install()— 19 callers, 4 callees - new:
_copy_skill_file()— 12 callers, 6 callees - new:
gemini_install()— 10 callers, 7 callees - new:
claude_uninstall()— 17 callers, 4 callees - new:
_project_uninstall()— 5 callers, 13 callees - …and 16 more — each is listed as a finding
Verification — 842 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 842 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
293 of 293 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safetytests/test_cross_repo_member_calls.py— full-run-safetytests/test_cross_repo_shared_types.py— full-run-safety- … and 243 more
non-code file(s) changed (
.github/workflows/ci.yml,.gitignore,CHANGELOG.md,README.md,graphify/omp/bun.lock…) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
.github/workflows/ci.yml,CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Formal verification
No difference found (not proven): No behavior difference found in \_is\_cwd\_relative (not a proof).
The verifier ran both versions of \_is\_cwd\_relative on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_hook\_guard (not a proof).
The verifier ran both versions of \_run\_hook\_guard on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).
The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).
The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
· 24 more finding(s) on lines outside this diff (see the check run).
Claude PreToolUse additionalContext parity: run the guard in tool_call as before (strict deny path unchanged), but deliver the nudge by appending it to that call's tool_result content instead of accumulating a deduped context-transform message. Every qualifying call now carries its own guidance inline, persisted across compaction like any tool output; the before_agent_start reset now only clears pending deliveries.
Copilot review finding: the README still described the replaced deduplicated context message. Describe the actual behavior: each qualifying call's guidance is appended to that call's persisted tool result.
Upstream OMP moved isReadableUrlPath from
@oh-my-pi/pi-coding-agent/tools/path-utils to @oh-my-pi/pi-tui/tools/read.
At the 18.2.2 devDependency floor path-utils still re-exported it, so CI
stayed green, but every OMP install past that point (18.2.6 is current)
throws SyntaxError: Export named 'isReadableUrlPath' not found before the
extension's api.on("tool_call", ...) ever registers, silently disabling
the hook-guard nudge.
Import it from @oh-my-pi/pi-tui/tools/read directly, raise the
@oh-my-pi/pi-coding-agent floor to ^18.2.6 (the first version pinned by
this fix where the export lives at its new home), and declare
@oh-my-pi/pi-tui ^18.2.6 as its own devDependency since the bridge now
imports from it directly rather than relying on it as an undeclared
transitive dependency of pi-coding-agent.
bun install --frozen-lockfile in the omp-bridge job pins an exact known-good OMP version, so CI can never observe a future upstream symbol move (this is exactly how the pi-tui export move shipped unnoticed: the lockfile pinned 18.2.2, where isReadableUrlPath still lived in pi-coding-agent). Add a step that installs whatever OMP publishes today and re-runs the suite against it, continue-on-error like the security-scan job's advisory checks, so an unrelated upstream release cannot block a merge while still surfacing a real break.
…oject guard _is_cwd_relative answered "no root and no drive" to decide whether a file_path/path value is cwd-anchored. A URL is rootless and driveless by that exact same test (https://x, myscheme://x, and a bare www.host/path all have no root/drive), so it short-circuited straight to "in project" and the containment check below never ran. - _normalize_hook_path: trims/de-quotes input and strips a leading file:// scheme to the local path it names (OMP's own pipeline resolves file:// to a local path, not an external URL -- it must still nudge). - _is_foreign_url_scheme: rejects any other whole-value scheme:// prefix. Deliberately does not enumerate a harness's internal-scheme vocabulary (local://, artifact://, ...) -- an unrecognized scheme is simply treated as not-a-local-source-file. - _has_embedded_url_scheme_segment: catches the same URL after an upstream host has already glued it onto an absolute prefix (<root>/local:/x), which looks exactly like a real file to the (unchanged) containment check. - _is_external_www_target: a bare www.host/path carries no :// at all, so it is handled separately, existence-gated to mirror OMP's own "an existing local path wins over URL" precedence. Containment, the extension allow-list, and staleness/strict-deny logic are untouched.
- isRemote gets a file:// carve-out: OMP's own resolveReadPathAsync
already reduces file:// to the real local path (its
strictExternalUrlRe deliberately excludes file), so routing it
through yields the correct absolute path instead of silently
dropping every file:// read. isInternalUrlPath/isReadableUrlPath
keep earning their place -- resolveReadPathAsync still pre-resolves
a rootless *foreign*-scheme value onto cwd into something that reads
as a real in-project file, so the bridge still must not hand it
those.
- resolveReadPath -> resolveReadPathAsync: same variant order/winner
semantics, non-blocking probes instead of statSync per candidate on
the tool_call hot path (read.ts itself already uses the async
variant).
- Drop the explicit { splitter: parseSearchPath } argument to
expandDelimitedPathEntries -- already its default.
Adds one bridge-level regression test: an in-project file:// target
still nudges, an out-of-project one still stays silent.
The OMP host loads graphify/omp/index.ts from its installed site-packages location, which ships no node_modules (see package.json's "files" list). The host provides @oh-my-pi/pi-coding-agent to legacy extensions but never @oh-my-pi/pi-tui, so the bridge's `isReadableUrlPath` import from @oh-my-pi/pi-tui/tools/read failed at runtime with `Cannot find package '@oh-my-pi/pi-tui'`, disabling the extension entirely. Vendor the two-line isReadableUrlPath predicate locally instead of importing it, drop the now-unused @oh-my-pi/pi-tui devDependency and regenerate bun.lock, and update the CI advisory step that used to bump both packages. Add a regression test asserting index.ts only imports @oh-my-pi/pi-coding-agent and node: builtins, so a future reintroduction of an unresolvable package fails the suite instead of only the host.
…er host copies Two @oh-my-pi/pi-coding-agent copies are reachable on this machine (18.1.17 and 18.2.6) and we cannot determine from outside the host which one it binds. Auditing every bridge import against the older 18.1.17 copy found resolveReadPathAsync missing -- it was only added in a later release as a non-blocking-probe performance optimization (per its own upstream docstring: "identical variant order and winner semantics, but non-blocking probes"), not a correctness fix. Revert to the synchronous resolveReadPath, present in both copies, so the extension loads regardless of which copy the host binds. A marginally faster extension that fails to load is worth nothing. Strengthen the import allow-list test with a version-floor check: every named import the bridge takes from @oh-my-pi/pi-coding-agent/tools/path-utils must be exported by the oldest host-provided copy. Parses the real import list out of index.ts rather than hardcoding it, so a future added symbol is checked automatically. Skips cleanly when that copy is not present on the machine running the suite.
…ct file _normalize_hook_path unconditionally reduced any file:// value to its path component, discarding the authority. file://evil.com/<in-project path> then classified identically to file:///<in-project path>: the same absolute path, silently treated as local. RFC 8089 / Node's url.fileURLToPath (ERR_INVALID_FILE_URL_HOST) say a file:// URL is local only when its authority is empty or localhost. Reduce only those two forms; leave any other authority intact so _is_foreign_url_scheme classifies it as the remote reference it is.
isRemote's file:// carve-out matched the scheme alone: any authority was treated as local, so file://evil.com/<in-project path> reached resolveReadPath, resolved to the real local path, and reached the guard as if it named a file in the project. Mirror Node's own url.fileURLToPath (ERR_INVALID_FILE_URL_HOST): only an empty or localhost authority is local. Any other authority now falls through to the generic scheme:// remote check, same as it would for https:// or any other foreign scheme.
The advisory job installed @oh-my-pi/pi-coding-agent@latest and ran the bridge suite against it on every CI run -- mutable, unpinned third-party code executing unconditionally. Replace the install-and- run with a download-only check: fetch the package with install scripts disabled, then statically compare index.ts's imported path-utils symbols against that file's declared exports as plain text. The downloaded package is never imported or executed, but the same upstream symbol move (isReadableUrlPath, pi-coding-agent -> pi-tui) that motivated this step still surfaces.
22caf5e to
ed70c8b
Compare
…xt through native hooks Squashed from origin/feat/omp-native-integration (96d155e) onto upstream/v8.
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 2 advisory finding(s) below merit a look before merge.
Formal verification. 4 change(s) tested, no difference found (not proven).
Graphify review — findings
Adds first-class Oh My Pi (OMP) support via a native guard extension: graphify omp install delegates to omp plugin install on the shipped graphify/omp/ package directory (linked, not vendored as npm/shell shims), graphify omp path prints that directory, and uninstall is done through omp plugin uninstall graphify-omp. The extension gates native read/glob/grep/search-bash calls through the installed graphify hook-guard CLI, translating policy denials into OMP block/reason and appending per-call guidance to tool results (with local-only path inspection, size/time caps, fail-open on error, and cancellation on session boundaries); it checks ctx.isProjectTrusted() but that method is currently always-true, so it's not an enforced sandbox. Adds a CI omp-bridge job that runs the tests/omp.test.ts suite against the pinned host plus real Python CLI, and an advisory static check that flags when @oh-my-pi/pi-coding-agent@latest stops exporting the path-utils symbols index.ts imports.
Worth a look
- Glob tool contract drops the user glob pattern —
graphify/omp/index.ts:125· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Test mutates global PATH without restoring it —
tests/omp.test.ts· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 869 functions depend on the 533 functions this change touches.
Health — this change adds coupling hotspots:
- new:
main()— 98 callers, 3 callees - new:
dispatch_command()— 2 callers, 125 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
install()— 13 callers, 11 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
gemini_install()— 11 callers, 7 callees - new:
claude_install()— 19 callers, 4 callees - new:
_copy_skill_file()— 12 callers, 6 callees - …and 17 more — each is listed as a finding
Verification — 869 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 869 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
303 of 303 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_cargo_missing_manifest.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_cobol_extractor.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safety- … and 253 more
non-code file(s) changed (
.github/workflows/ci.yml,.gitignore,CHANGELOG.md,README.md,graphify/omp/bun.lock…) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
.github/workflows/ci.yml,CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Formal verification
No difference found (not proven): No behavior difference found in \_is\_cwd\_relative (not a proof).
The verifier ran both versions of \_is\_cwd\_relative on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_hook\_guard (not a proof).
The verifier ran both versions of \_run\_hook\_guard on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in dispatch\_install\_cli (not a proof).
The verifier ran both versions of dispatch\_install\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_run\_cli (not a proof).
The verifier ran both versions of \_run\_cli on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
· 25 more finding(s) on lines outside this diff (see the check run).
Claude Code wraps PreToolUse additionalContext in a system reminder that names the hook; the OMP bridge appended the bare guidance after the tool output, where it read as text the tool printed and agents discounted it as untrusted. Lead each qualifying result with one <system-reminder source="graphify"> block instead, the shape OMP's own per-tool TTSR reminders use, and keep the tool output unchanged after it.
…xt through native hooks Squashed from origin/feat/omp-native-integration (25e46e5) onto upstream/v8.
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 2 advisory finding(s) below merit a look before merge.
Graphify review — findings
Adds a native Oh My Pi guard integration installed via graphify omp install, which links the shipped graphify/omp package through OMP's own omp plugin install route and runs the existing hook-guard read|search policy before native read/glob/grep/search-bash calls, injecting per-call guidance as a leading <system-reminder source="graphify"> block on each tool result and honoring GRAPHIFY_HOOK_STRICT=1 denials without ever building a graph. Hooks only inspect local filesystem targets, require an absolute-PATH graphify outside the project with no fallback, and enforce 256 KiB input / 64 KiB output / two-second budgets that fail open; omp is now silenced from the stale-graph warning path and graphify omp path prints the package directory for manual linking. Adds an omp-bridge CI job that runs the bridge tests against the pinned OMP version as the blocking gate plus an advisory, non-executing static check that a @latest download still exports the path-utils symbols index.ts imports.
Worth a look
- Read tool guard ignores file_path and inspects cwd —
graphify/omp/index.ts:123· Escalate · high- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Test mutates process.env.PATH without restoration —
tests/omp.test.ts· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 871 functions depend on the 535 functions this change touches.
Health — this change adds coupling hotspots:
- new:
main()— 98 callers, 3 callees - new:
dispatch_command()— 2 callers, 125 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
install()— 13 callers, 11 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
gemini_install()— 11 callers, 7 callees - new:
claude_install()— 19 callers, 4 callees - new:
_copy_skill_file()— 12 callers, 6 callees - …and 17 more — each is listed as a finding
Verification — 871 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 871 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
303 of 303 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_cargo_missing_manifest.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_cobol_extractor.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safety- … and 253 more
non-code file(s) changed (
.github/workflows/ci.yml,.gitignore,CHANGELOG.md,README.md,graphify/omp/bun.lock…) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
.github/workflows/ci.yml,CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
· 25 more finding(s) on lines outside this diff (see the check run).
OMP reads its agent skills from ~/.omp/agent/skills, not the legacy ~/.pi path the pi platform writes, so the installer had no OMP target even though 'omp' was already reserved in the CLI command set. Adds the omp platform reusing pi's skill bundle (OMP mirrors pi's agent layout), the pi-twin 'graphify omp' subcommand, and the project-scope groups; tests lock destinations, bundle reuse, install/uninstall, and uninstall_all pi parity.
…staller The merged omp-platform branch added a pi-twin 'graphify omp [install|uninstall]' dispatch branch, unreachable behind the bridge PR's 'graphify omp [path|install]' plugin branch. The subcommand surface belongs to the plugin installer: drop the duplicate, and carry the user-scope skill removal in uninstall_all (like amp/agents) since omp has no subcommand twin.
…xt through native hooks Squashed from origin/feat/omp-native-integration (0fd7486) onto upstream/v8.
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 5 advisory finding(s) below merit a look before merge.
Graphify review — findings
Adds an Oh My Pi native guard integration: graphify omp install links the shipped graphify/omp package (with package.json declaring omp.extensions) via OMP's omp plugin install, and graphify omp path prints that directory for manual linking. The extension runs the installed graphify hook-guard before native read/glob/grep/search bash calls, turning policy denials into OMP block/reason and appending per-call guidance as a leading <system-reminder source="graphify"> block on tool results, inspecting only local filesystem targets and failing open on missing commands, bad output, or the 2s/256 KiB/64 KiB budgets. Adds an omp-bridge CI job that runs the bun bridge tests against the frozen pinned OMP plus a non-blocking advisory step that statically checks whether @latest still exports the path-utils symbols index.ts imports; index.ts now vendors isReadableUrlPath locally rather than importing it.
Worth a look
- Glob hook payload overwrites the user's glob pattern with a resolved path —
graphify/omp/index.ts:132· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- New CI job executes mutable third-party GitHub Actions —
.github/workflows/ci.yml:88· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- omp platform config duplicates 'pi' skill_dst under project scope but user-scope uses ~/.omp/agent/skills —
graphify/install.py:1934· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Relative GRAPHIFY_TEST_CLI is shadowed by the fixture —
tests/omp.test.ts:10· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
- Test mutates global PATH without restoring it —
tests/omp.test.ts· Escalate · medium- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 889 functions depend on the 553 functions this change touches.
Health — this change adds coupling hotspots:
- new:
main()— 99 callers, 3 callees - new:
dispatch_command()— 2 callers, 125 callees - new:
dispatch_install_cli()— 6 callers, 31 callees - new:
install()— 13 callers, 11 callees - new:
codebuddy_install()— 20 callers, 5 callees - new:
gemini_install()— 11 callers, 7 callees - new:
claude_install()— 19 callers, 4 callees - new:
_copy_skill_file()— 12 callers, 6 callees - …and 17 more — each is listed as a finding
Verification — 889 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 889 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
304 of 304 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/omp.test.ts— impact, changed-test, full-run-safetytests/test_affected_cli.py— impact, full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— impact, full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— impact, full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_cargo_missing_manifest.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— impact, full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_cobol_extractor.py— full-run-safetytests/test_codebuddy.py— impact, full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safety- … and 254 more
non-code file(s) changed (
.github/workflows/ci.yml,.gitignore,CHANGELOG.md,README.md,graphify/omp/bun.lock…) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
.github/workflows/ci.yml,CHANGELOG.md,README.md,graphify/omp/package.json,pyproject.toml) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
· 25 more finding(s) on lines outside this diff (see the check run).
This branch has not been deployed
1 participant
Summary
Bridge Graphify's existing hook-guard policy into OMP's native tool_call/context hooks, including strict denials. Ship the extension inside the Python distribution and register it through OMP's plugin manager with graphify omp install.
<system-reminder source="graphify">block, the way Claude Code labels PreToolUse additionalContext (and OMP labels its own per-tool TTSR reminders), clearing pending deliveries across runs and session navigation.Related work
Related to #1937, which adds OMP nudges on tool results. This patch targets a different runtime contract: reuse hook-guard before execution so strict denials, native grep, path selectors, and policy context actually work. It does not duplicate hardcoded tool-result guidance. There is installation/documentation overlap; happy to coordinate the landing order.
Verification
No existing project index was built or changed. OMP's isProjectTrusted() currently always returns true and is not a sandbox; docs say so. Tested host: OMP18.1.17 integration SHA6aef0e8ad51b3bc5ea7a5f2a255c3d48e4c5af72, not a promise about all published builds with that version. The installed executable policy excludes a project-local .venv; docs recommend uv tool/pipx.
AI-assisted implementation, independently reviewed for contracts, hostile input, test honesty and unnecessary complexity.