feat(extract): Add GDScript extractor by AWCostabile · Pull Request #3750 · Graphify-Labs/graphify

AWCostabile · 2026-09-22T14:29:33Z

What does this PR do?

This adds support for GDScript, the scripting language in Godot 4 (hopefully to assist those in indie game development). I use graphify on a single Godot project and the .gd files were being skipped, so most of the codebase was missing from the graph 😢!

While GDScript doesn't have import statements, a script can refers to another by:
  a) extending it
  b) calling preload("res://path.gd")
  c) using one of the autoload singletons listed in project.godot
  d) using a global class_name
So the extractor is built around these four mechanisms.

For each script it creates a node for the file (the label is the class_name if there is one, otherwise the autoload name, otherwise the filename), and nodes for the functions, signals, constants, enums and inner classes in it. Methods of an inner class are labelled .name() with a method link back to the class, the same way the Zig extractor handles struct methods.

What it records between scripts:

extends X becomes an inherits relation pointing at the parent script. Built-in engine classes such as Node don't have a script file, so nothing is emitted for them.
preload() and load() of a .gd path are recorded as imports of that file. This is the important one. Scripts that don't declare a class_name can only be reached this way.
signal.connect(callback), signal.emit() and emit_signal("name") are recorded as uses from the enclosing function to the signal. The callback handed to connect is also noted as a references relation.
Unqualified calls are looked up in the current file first and then up the extends chain.
Autoload.method(), Alias.method() where Alias is a preload constant, and ClassName.method() are looked up in the script they name. .new() on any of those is recorded as references against the script.
A comment containing something like movement.md §2.14 links the enclosing function (or the file) to a section node for that page as references. The project I work on does this in almost every file, so it was worth the few lines. If a later comment in the same file just says §2.15, it is attached to the last page that file mentioned and marked INFERRED. Everything else above is EXTRACTED.

Anything that can't be resolved with those rules (engine methods like add_child, calls on $Node, calls on local variables) is counted in unresolved_calls and left out. I did try passing them through as raw_calls so the shared cross-file resolver could match them by name, but on real code that mostly connected _ready and _process between scripts that have nothing to do with each other. Leaving them out gave a much cleaner graph as far as I see it.

The autoload table, the class_name index and the per-file function and signal names are read once per project.godot and cached for the lifetime of the process. Autoloads written as uid:// are resolved through .gd.uid files which Godot generates next to each script. If a given script isn't inside a Godot project at all it still gets extracted (just without cross-file resolution).

*On ids: functions, signals and classes use make_id(stem, name) so the existing remap pass handles them. Constants and enums include a kind segment (make_id(stem, "const", name)). I had a fixture with const Helper and func helper() in the same file and they merged into one node, because make_id lowercases.

Type of change

New feature
Tests or CI
Documentation

Files changed

graphify/extractors/gdscript.py is new.
graphify/extractors/__init__.py, graphify/extract.py (re-export, _DISPATCH, _LANG_FAMILY_BY_EXT) and graphify/detect.py (CODE_EXTENSIONS) register the suffix.
- The watcher's extension set is derived from CODE_EXTENSIONS, so it didn't need a change.
pyproject.toml and uv.lock add tree-sitter-gdscript>=6.1,<7. It's a small wheel on PyPI and works with the tree-sitter 0.25 already pinned. tree-sitter-language-pack would also do it but that pulls in 160 grammars.
- If you'd rather have this as an optional extra like [dm], that's fine with me (the tests already skip when the module is absent).
Tests are in tests/test_languages.py and tests/test_extractors_registry.py, with the fixture at tests/fixtures/sample.gd.
Added the extension in the grammar row of README.md, and the count was bumped to 38.
- Also added a status row to graphify/extractors/MIGRATION.md.

How was this tested?

uv sync
uv run pytest tests/test_languages.py tests/test_extractors_registry.py -k gdscript -q
  6 passed
uv run pytest tests/ -q        (Windows 11, CPython 3.13)
  23 failed, 5676 passed, 138 skipped
uv run ruff check graphify/extractors/gdscript.py tests/test_languages.py
  All checks passed!

The 23 failures are already there on v8 at 20a20d3 before my changes, same tests by name. They look Windows-specific to me: the read-only file case in test_atomic_writes, the symlink cases in test_non_regular_files, test_ollama_retry_cap, the install and uninstall scope tests, two in test_hooks, two in test_watch, one terraform path test, but none of them involve .gd.

The new tests set up a small Godot project in tmp_path with two autoloads (one declared as uid://), a script that has no class_name and can only be reached by preload, and a helper. They check the imports, inheritance and calls across three scripts, the signal wiring, the autoload label, a script outside any project, and that .gd is routed to this extractor only.

I also ran it on the project it was written for, which has 174 scripts (174 file nodes, 2842 functions and 281 imports). Grepping the source finds 288 preload/load targets. Six of the seven that don't show up are examples inside doc comments, and the last is a pair of scripts that preload each other, which the undirected graph keeps as a single connection. There are 477 citation links to 152 section nodes, nothing dangling and no absolute paths in any id. graphify path between two scripts that only connect through a third returns the two-hop chain.

Graphify-specific checklist

I added or updated tests for behavior changes.
I updated documentation or confirmed that no documentation is needed.
I updated generated skill artifacts when changing their source fragments.
- *Nothing to update, the skill files don't list languages.
I considered compatibility across supported Python versions. No syntax newer than 3.10.
I confirmed that no API keys, generated graph data, or local-only files are included.

Additional Information

I used imports for preloads rather than imports_from, because a preload brings in the whole script and the relation points at the file node. If imports_from is what you'd expect there, that's a simple enough change. The doc citation nodes have file_type: "doc" and source_file set to the markdown page (resolved under docs/ when that directory exists). I didn't find an existing node type that fit so I used section, but I am happy to rename it if you prefer.

PS: Cheers for a great project 👍

Nodes: the script (its class_name, or its autoload name, or the file name), func, signal, const, enum and inner class. Edges: extends -> inherits to the parent script; preload()/load() of a .gd -> imports; signal connect/emit -> uses (plus the connect callback as a reference); calls resolved through the same file, the extends chain, the project's [autoload] table, preload const aliases and the class_name index - all EXTRACTED, because each names its target in source; a call none of those resolves is counted, never guessed, and never handed to the shared name-matching pass. A <page>.md §N.N in a comment becomes a references edge to a section node of that page; a bare §N.N inherits the last page the file named (INFERRED). The project index is built once per project.godot root and cached per process. Registered in the dispatch, the language-family table, the extractor registry and detect's code set; grammar tree-sitter-gdscript 6.x. Version carries the +gms1 local label so the fork build is observable from --version.

A fixture script exercising every declaration shape, extracted inside a tmp_path Godot project (autoloads by res:// and uid://, a preload-only law script, a helper); inheritance and ancestor-call resolution across three scripts; the autoload label; a script outside any project; and the dispatch assertion that no other grammar reaches a .gd file.

graphify-labs

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.

Graphify review — findings

Adds a GDScript extractor and wires .gd through the pipeline: extract_gdscript in gdscript.py treats one file as one class and emits its func/signal/const/enum/inner-class members, with edges for extends inheritance, preload/load imports, signal connect/emit/emit_signal uses, and call resolution through same-file/ancestor functions plus a per-project [autoload], class_name, and uid:// index (cached per project.godot root). Registers .gd in CODE_EXTENSIONS, the language-family map, and the extractor dispatch/registry so detection and dispatch pick it up; calls resolved by none of the project mechanisms are counted as unresolved_calls rather than fed to the shared name-matcher, and files outside a Godot project still extract with only same-file resolution. Bumps the README grammar count to 38 and adds .gd to the listed extensions.

Worth a look

Path traversal via unvalidated citation page in doc section resolution — graphify/extractors/gdscript.py · Escalate · medium
- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Failed project.godot read leaves _uids uninitialized — graphify/extractors/gdscript.py:82 · Escalate · medium
- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
Bare citations in a comment attach to the last explicit page, not the nearest preceding page — graphify/extractors/gdscript.py · 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 — 3088 functions depend on the 1025 functions this change touches.

Health — this change adds coupling hotspots:

new: extract() — 655 callers, 45 callees
new: _rebuild_code() — 142 callers, 54 callees
new: detect() — 112 callers, 15 callees
new: _extract_generic() — 18 callers, 29 callees
new: save_manifest() — 40 callers, 11 callees
new: extract_js() — 87 callers, 4 callees
new: extract_files_direct() — 17 callers, 20 callees
new: extract_xaml() — 19 callers, 17 callees
…and 61 more — each is listed as a finding

Verification — 3088 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: 2837 function(s) in the blast radius were not formally verified this run

Test selection

290 of 290 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/test_affected_cli.py — full-run-safety
tests/test_affected_member_seed.py — full-run-safety
tests/test_agents_platform.py — full-run-safety
tests/test_analyze.py — full-run-safety
tests/test_anthropic_custom_endpoint.py — full-run-safety
tests/test_antigravity_install.py — full-run-safety
tests/test_apm_fallback_version.py — full-run-safety
tests/test_architecture_doc.py — full-run-safety
tests/test_astro_extraction.py — impact, full-run-safety
tests/test_astro_import_ids.py — impact, full-run-safety
tests/test_atomic_canvas_export.py — full-run-safety
tests/test_atomic_version_stamp.py — full-run-safety
tests/test_atomic_writes.py — impact, full-run-safety
tests/test_backend_env_isolation.py — full-run-safety
tests/test_backend_extras.py — full-run-safety
tests/test_benchmark.py — full-run-safety
tests/test_benchmark_raw_graph.py — full-run-safety
tests/test_build.py — impact, full-run-safety
tests/test_build_merge_dedup_scope.py — full-run-safety
tests/test_build_merge_hyperedges_and_prune.py — full-run-safety
tests/test_build_merge_shrink_guard.py — full-run-safety
tests/test_builtin_global_type_refs.py — impact, full-run-safety
tests/test_cache.py — full-run-safety
tests/test_callflow_html.py — full-run-safety
tests/test_cargo_introspect.py — full-run-safety
tests/test_carried_hyperedge_remap.py — full-run-safety
tests/test_case_sensitive_resolution.py — impact, full-run-safety
tests/test_charmap_encoding.py — impact, full-run-safety
tests/test_chunking.py — impact, full-run-safety
tests/test_cjs_module_extension.py — impact, full-run-safety
tests/test_claude_cli_backend.py — impact, full-run-safety
tests/test_claude_md.py — full-run-safety
tests/test_cli_broken_pipe.py — full-run-safety
tests/test_cli_export.py — full-run-safety
tests/test_cli_help.py — full-run-safety
tests/test_cluster.py — full-run-safety
tests/test_codebuddy.py — full-run-safety
tests/test_community_hub_labels.py — full-run-safety
tests/test_community_labels_skill.py — full-run-safety
tests/test_confidence.py — full-run-safety
tests/test_corrupt_graph_json.py — full-run-safety
tests/test_cpp_nested_and_cli.py — impact, full-run-safety
tests/test_cpp_objc_cross_file_calls.py — impact, full-run-safety
tests/test_cpp_preprocess.py — full-run-safety
tests/test_cross_extension_reexport_self_cycle.py — impact, full-run-safety
tests/test_cross_language_call_resolution.py — impact, full-run-safety
tests/test_cross_repo_external_call_guards.py — impact, full-run-safety
tests/test_cross_repo_member_calls.py — impact, full-run-safety
tests/test_cross_repo_shared_types.py — full-run-safety
tests/test_csharp_call_site_generic_args.py — impact, full-run-safety
… and 240 more

non-code file(s) changed (README.md, graphify/extractors/MIGRATION.md, pyproject.toml, tests/fixtures/sample.gd, uv.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 (README.md, graphify/extractors/MIGRATION.md, graphify/extractors/__init__.py, 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.

· 5 grounded finding(s) anchored inline below; 64 more finding(s) on lines outside this diff (see the check run).

graphify-labs · 2026-09-22T14:46:12Z

+    return out
+
+
+def extract_gdscript(path: Path) -> dict:


⚠️ Health regression — extract_gdscript()

fans out to 12 callees (efferent coupling); 7 callers depend on it (afferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-09-22T14:46:12Z

+    function_bodies: list[tuple[str, Any, Any]] = []   # (nid, body node, def node)
+    top_funcs: frozenset[str] = frozenset()
+
+    def declare_member(node, owner_nid: str, owner_stem: str, inner: bool) -> None:


⚠️ Health regression — declare_member()

fans out to 7 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-09-22T14:46:12Z

+                return _script_of(value, path, project) if value.endswith(".gd") else None
+        return None
+
+    def emit_extends(node, owner_nid: str) -> None:


⚠️ Health regression — emit_extends()

fans out to 6 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-09-22T14:46:12Z

+                    return _read_text(parts[1], source)
+        return None
+
+    def handle_attribute(node, caller_nid: str) -> None:


⚠️ Health regression — handle_attribute()

fans out to 11 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-09-22T14:46:12Z

+                return
+        unresolved_calls.append({"caller_nid": caller_nid, "callee": f"{'.'.join(names)}.{method}" if names else method, "line": line})
+
+    def handle_call(node, caller_nid: str) -> None:


⚠️ Health regression — handle_call()

fans out to 6 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

AWCostabile · 2026-09-22T14:50:51Z

Ahh, I jus saw #1929 after opening, which is pretty thorough 😓... While this PR covers the GDScript half of Godot Support on the current v8 with tests, the scene-file extractor from #1929 probably makes that a more valueable PR overall, that this could only really build upon (or form a base to be built upon)... Happy to sit by if that's preferred, but will leave open until checked.

Addresses the three advisory findings from the automated review: - the project index initialises its uid table before anything can fail, so a project.godot that cannot be read still leaves an index that answers None for a uid:// instead of raising AttributeError; - a cited documentation page only rewrites the recorded path when it resolves to a real file INSIDE the project root (or its docs/), and a page carrying '..' segments is never looked up at all — the caller already hands a basename, but the resolver now enforces containment itself; - citations are walked in the order they appear in a comment, so a bare section attaches to the page named before it and never to one named later on the same line. One test per finding.

AWCostabile · 2026-09-22T15:06:51Z

Update: I pushed b569873 to address the three advisory findings from the bot review!

uid table init: the project index now sets up its uid:// map before anything can fail, so an unreadable project.godot returns "None" instead of raising.
citation page containment: a cited page only rewrites the recorded path if it resolves to a real file inside the project root (or docs/), and anything with .. in it is never looked up. The caller was already passing a basename, but the resolver enforces it itself now.
bare § ordering: citations are walked in the order they appear in the comment, so a bare section takes the page named before it, never one named later on the same line.

Added one test per finding, and the full suite runs locally with 5679 passes and the same 23 Windows-only failures as a clean v8 checkout. The five fan-out notes are the usual for a new extractor module (the Zig/Rust ones score about the same), so I've left those as-is.

graphify-labs

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 GDScript extractor for Godot 4 .gd files, wiring extract_gdscript into the dispatch table, the dotnet-style language-family map, the extractor registry, and CODE_EXTENSIONS. The extractor treats one file as one class and emits Godot-aware edges: extends → inherits to the parent script, preload/load → imports, signal connect/emit → uses (plus references to a callback), and calls resolved through the project's [autoload] table, preload const aliases, and class_name index — with unresolvable bare calls counted as unresolved_calls rather than handed to the name-matching resolver. Builds a per-project.godot index (autoloads, class_name→script, uid://→script, per-script func names) cached per process, and falls back to same-file-only resolution for scripts outside any Godot project.

Worth a look

_file_index returns stale declarations after a script is edited — graphify/extractors/gdscript.py:134 · Escalate · medium
- agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review
GDScript res:// references can escape the project root — graphify/extractors/gdscript.py:97 · 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 — 3092 functions depend on the 1029 functions this change touches.

Health — this change adds coupling hotspots:

new: extract() — 655 callers, 45 callees
new: _rebuild_code() — 142 callers, 54 callees
new: detect() — 112 callers, 15 callees
new: _extract_generic() — 18 callers, 29 callees
new: save_manifest() — 40 callers, 11 callees
new: extract_js() — 87 callers, 4 callees
new: extract_files_direct() — 17 callers, 20 callees
new: extract_xaml() — 19 callers, 17 callees
…and 61 more — each is listed as a finding

Verification — 3092 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: 2841 function(s) in the blast radius were not formally verified this run

Test selection

290 of 290 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/test_affected_cli.py — full-run-safety
tests/test_affected_member_seed.py — full-run-safety
tests/test_agents_platform.py — full-run-safety
tests/test_analyze.py — full-run-safety
tests/test_anthropic_custom_endpoint.py — full-run-safety
tests/test_antigravity_install.py — full-run-safety
tests/test_apm_fallback_version.py — full-run-safety
tests/test_architecture_doc.py — full-run-safety
tests/test_astro_extraction.py — impact, full-run-safety
tests/test_astro_import_ids.py — impact, full-run-safety
tests/test_atomic_canvas_export.py — full-run-safety
tests/test_atomic_version_stamp.py — full-run-safety
tests/test_atomic_writes.py — impact, full-run-safety
tests/test_backend_env_isolation.py — full-run-safety
tests/test_backend_extras.py — full-run-safety
tests/test_benchmark.py — full-run-safety
tests/test_benchmark_raw_graph.py — full-run-safety
tests/test_build.py — impact, full-run-safety
tests/test_build_merge_dedup_scope.py — full-run-safety
tests/test_build_merge_hyperedges_and_prune.py — full-run-safety
tests/test_build_merge_shrink_guard.py — full-run-safety
tests/test_builtin_global_type_refs.py — impact, full-run-safety
tests/test_cache.py — full-run-safety
tests/test_callflow_html.py — full-run-safety
tests/test_cargo_introspect.py — full-run-safety
tests/test_carried_hyperedge_remap.py — full-run-safety
tests/test_case_sensitive_resolution.py — impact, full-run-safety
tests/test_charmap_encoding.py — impact, full-run-safety
tests/test_chunking.py — impact, full-run-safety
tests/test_cjs_module_extension.py — impact, full-run-safety
tests/test_claude_cli_backend.py — impact, full-run-safety
tests/test_claude_md.py — full-run-safety
tests/test_cli_broken_pipe.py — full-run-safety
tests/test_cli_export.py — full-run-safety
tests/test_cli_help.py — full-run-safety
tests/test_cluster.py — full-run-safety
tests/test_codebuddy.py — full-run-safety
tests/test_community_hub_labels.py — full-run-safety
tests/test_community_labels_skill.py — full-run-safety
tests/test_confidence.py — full-run-safety
tests/test_corrupt_graph_json.py — full-run-safety
tests/test_cpp_nested_and_cli.py — impact, full-run-safety
tests/test_cpp_objc_cross_file_calls.py — impact, full-run-safety
tests/test_cpp_preprocess.py — full-run-safety
tests/test_cross_extension_reexport_self_cycle.py — impact, full-run-safety
tests/test_cross_language_call_resolution.py — impact, full-run-safety
tests/test_cross_repo_external_call_guards.py — impact, full-run-safety
tests/test_cross_repo_member_calls.py — impact, full-run-safety
tests/test_cross_repo_shared_types.py — full-run-safety
tests/test_csharp_call_site_generic_args.py — impact, full-run-safety
… and 240 more

non-code file(s) changed (README.md, graphify/extractors/MIGRATION.md, pyproject.toml, tests/fixtures/sample.gd, uv.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 (README.md, graphify/extractors/MIGRATION.md, graphify/extractors/__init__.py, 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.

· 5 grounded finding(s) anchored inline below; 64 more finding(s) on lines outside this diff (see the check run).

graphify-labs · 2026-09-22T15:32:03Z

+    return out
+
+
+def extract_gdscript(path: Path) -> dict:


⚠️ Health regression — extract_gdscript()

fans out to 12 callees (efferent coupling); 9 callers depend on it (afferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-09-22T15:32:03Z

+    function_bodies: list[tuple[str, Any, Any]] = []   # (nid, body node, def node)
+    top_funcs: frozenset[str] = frozenset()
+
+    def declare_member(node, owner_nid: str, owner_stem: str, inner: bool) -> None:


⚠️ Health regression — declare_member()

fans out to 7 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-09-22T15:32:03Z

+                return _script_of(value, path, project) if value.endswith(".gd") else None
+        return None
+
+    def emit_extends(node, owner_nid: str) -> None:


⚠️ Health regression — emit_extends()

fans out to 6 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-09-22T15:32:03Z

+                    return _read_text(parts[1], source)
+        return None
+
+    def handle_attribute(node, caller_nid: str) -> None:


⚠️ Health regression — handle_attribute()

fans out to 11 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

graphify-labs · 2026-09-22T15:32:03Z

+                return
+        unresolved_calls.append({"caller_nid": caller_nid, "callee": f"{'.'.join(names)}.{method}" if names else method, "line": line})
+
+    def handle_call(node, caller_nid: str) -> None:


⚠️ Health regression — handle_call()

fans out to 6 callees (efferent coupling).

Grounded coupling-delta finding (deterministic), not an LLM guess.

anthonycostabile added 4 commits September 22, 2026 23:16

docs: list GDScript among the supported grammars

693d46c

chore: strip +gms1 from version number for PR to original repo

a1561e3

graphify-labs Bot reviewed Sep 22, 2026

View reviewed changes

Uh oh!

Conversation

AWCostabile commented Sep 22, 2026

What does this PR do?

What it records between scripts:

Type of change

Files changed

How was this tested?

Graphify-specific checklist

Additional Information

Uh oh!

graphify-labs Bot left a comment

Choose a reason for hiding this comment

Graphify review — findings

Impact & health

Graphify review

Gate & verification

graphify gate

Test selection

Test selection

Uh oh!

graphify-labs Bot Sep 22, 2026

Choose a reason for hiding this comment

Uh oh!

graphify-labs Bot Sep 22, 2026

Choose a reason for hiding this comment

Uh oh!

graphify-labs Bot Sep 22, 2026

Choose a reason for hiding this comment

Uh oh!

graphify-labs Bot Sep 22, 2026

Choose a reason for hiding this comment

Uh oh!

graphify-labs Bot Sep 22, 2026

Choose a reason for hiding this comment

Uh oh!

AWCostabile commented Sep 22, 2026

Uh oh!

AWCostabile commented Sep 22, 2026

Uh oh!

graphify-labs Bot left a comment

Choose a reason for hiding this comment

Graphify review — findings

Impact & health

Graphify review

Gate & verification

graphify gate

Test selection

Test selection

Uh oh!

graphify-labs Bot Sep 22, 2026

Choose a reason for hiding this comment

Uh oh!

graphify-labs Bot Sep 22, 2026

Choose a reason for hiding this comment

Uh oh!

graphify-labs Bot Sep 22, 2026

Choose a reason for hiding this comment

Uh oh!

graphify-labs Bot Sep 22, 2026

Choose a reason for hiding this comment

Uh oh!

graphify-labs Bot Sep 22, 2026

Choose a reason for hiding this comment

Uh oh!

This branch has not been deployed

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants