@skill-map/spec@0.59.0

Minor Changes

• The vendor-neutral open-skills Provider (agent-skills, lens "Open Skills") gains an open-standard base reserved-name catalog under skill: a user skill shadowing a universal built-in like help/config is now flagged by core/name-reserved, and Antigravity inherits the base by manifest composition and appends its own verbs. Its skill frontmatter schema now enforces the open-standard name pattern/length and description length. Shared primitives renamed to a COMMONS_* vocabulary.

  User-facing

  With the Open Skills lens active, a skill you authored that shares a name with a built-in command (like help or config) now gets a warning, and skill names or descriptions that break the open-standard format (bad characters, too long) are flagged too.

@skill-map/spec@0.58.0

Minor Changes

• Bare sm in an empty folder now offers a getting-started menu: on an interactive terminal it asks whether to run the guided tutorial (sm tutorial) or drop a ready-to-explore example project (sm example), then dispatches the chosen verb. In a non-empty folder, or on a non-interactive stdin, it still prints a one-line hint and exits 2, now pointing at sm tutorial / sm example when the folder is empty and at sm init otherwise.

  User-facing

  Run sm in an empty folder and it now asks how you want to start: a guided tutorial, or a ready-made example project to explore. Pick one and it sets it up for you.

• New sm example verb: drops a ready-to-explore example project (the same wired harness the public demo renders) into an empty directory, so a new user can run sm scan then sm serve against a real connected graph without authoring files first. The payload is the single canonical fixtures/demo-scope/ fixture, shared with the web demo, and ships unscanned (no .skill-map/). Refuses a non-empty cwd unless --force.

  User-facing

  New sm example command: run it in an empty folder to drop a small ready-made project, then sm scan and sm serve to explore it as a live graph. The fastest way to try skill-map without setting up your own files first.

@skill-map/cli@0.65.0

Minor Changes

• The vendor-neutral open-skills Provider (agent-skills, lens "Open Skills") gains an open-standard base reserved-name catalog under skill: a user skill shadowing a universal built-in like help/config is now flagged by core/name-reserved, and Antigravity inherits the base by manifest composition and appends its own verbs. Its skill frontmatter schema now enforces the open-standard name pattern/length and description length. Shared primitives renamed to a COMMONS_* vocabulary.

  User-facing

  With the Open Skills lens active, a skill you authored that shares a name with a built-in command (like help or config) now gets a warning, and skill names or descriptions that break the open-standard format (bad characters, too long) are flagged too.

@skill-map/cli@0.64.1

Patch Changes

• Patch release of @skill-map/cli with no functional change, used to exercise the changesets version-packages PR and the end-to-end release pipeline.

@skill-map/cli@0.64.0

Minor Changes

• Bare sm in an empty folder now offers a getting-started menu: on an interactive terminal it asks whether to run the guided tutorial (sm tutorial) or drop a ready-to-explore example project (sm example), then dispatches the chosen verb. In a non-empty folder, or on a non-interactive stdin, it still prints a one-line hint and exits 2, now pointing at sm tutorial / sm example when the folder is empty and at sm init otherwise.

  User-facing

  Run sm in an empty folder and it now asks how you want to start: a guided tutorial, or a ready-made example project to explore. Pick one and it sets it up for you.

• New sm example verb: drops a ready-to-explore example project (the same wired harness the public demo renders) into an empty directory, so a new user can run sm scan then sm serve against a real connected graph without authoring files first. The payload is the single canonical fixtures/demo-scope/ fixture, shared with the web demo, and ships unscanned (no .skill-map/). Refuses a non-empty cwd unless --force.

  User-facing

  New sm example command: run it in an empty folder to drop a small ready-made project, then sm scan and sm serve to explore it as a live graph. The fastest way to try skill-map without setting up your own files first.

@skill-map/spec@0.57.0

Minor Changes

• The active provider lens no longer has an unlensed (permissive) state. A project with no marker now resolves to the universal markdown lens (never null, never persisted, so a later vendor marker still auto-detects) instead of running every provider at once. The Settings dropdown drops the dead (none) entry and keeps Markdown as a selectable neutral lens, and sm serve now re-scans under the chosen lens after a switch instead of re-detecting it from disk.

  User-facing

  A repo with no .claude/, .codex/, or .agents/ now opens in the Markdown view instead of mixing every platform together, with no warning. Pick Markdown anytime from Settings to see your files as plain markdown. The empty (none) option is gone.

• Removed the comingSoon provider flag: not-ready providers use stability: 'experimental', shipping disabled by default (not classified, auto-detected, or selectable until enabled). openai, antigravity, agent-skills are experimental; agent-skills is gated to its own lens (only core/markdown stays universal). Antigravity reuses the agent-skills classifier, dropping the kernel's cross-provider reservedNames lens-scope. sm tutorial --experimental offers them as destinations.

  User-facing

  The lens dropdown no longer shows "(coming soon)" rows. Not-ready providers (OpenAI Codex, Antigravity, Open Skills) are hidden until you enable them with sm plugins enable <id>; sm tutorial --experimental offers them as tutorial destinations.

@skill-map/cli@0.63.0

Minor Changes

• The active provider lens no longer has an unlensed (permissive) state. A project with no marker now resolves to the universal markdown lens (never null, never persisted, so a later vendor marker still auto-detects) instead of running every provider at once. The Settings dropdown drops the dead (none) entry and keeps Markdown as a selectable neutral lens, and sm serve now re-scans under the chosen lens after a switch instead of re-detecting it from disk.

  User-facing

  A repo with no .claude/, .codex/, or .agents/ now opens in the Markdown view instead of mixing every platform together, with no warning. Pick Markdown anytime from Settings to see your files as plain markdown. The empty (none) option is gone.

• Removed the comingSoon provider flag: not-ready providers use stability: 'experimental', shipping disabled by default (not classified, auto-detected, or selectable until enabled). openai, antigravity, agent-skills are experimental; agent-skills is gated to its own lens (only core/markdown stays universal). Antigravity reuses the agent-skills classifier, dropping the kernel's cross-provider reservedNames lens-scope. sm tutorial --experimental offers them as destinations.

  User-facing

  The lens dropdown no longer shows "(coming soon)" rows. Not-ready providers (OpenAI Codex, Antigravity, Open Skills) are hidden until you enable them with sm plugins enable <id>; sm tutorial --experimental offers them as tutorial destinations.

@skill-map/spec@0.56.1

Patch Changes

• The /api/branch map projection now keeps an edge when its RESOLVED target is a rendered node, not only when the raw authored target is. Trigger-style invokes / mentions links store the trigger (/cmd, @agent) in target and the real node path in resolvedTarget; the old filter matched the raw target alone, so every resolved trigger edge was dropped from the graph and the map showed only path-style references. Genuinely-broken links (no resolved node) stay excluded.

  User-facing

  The graph map again draws invokes and mentions arrows (a command running a skill, an agent referenced by name), not just plain file references. A recent change had hidden every resolved trigger edge from the map.

@skill-map/spec@0.56.0

Minor Changes

• Splits the scan cap into two knobs: scan.maxScan (corpus ceiling, default 50000) bounds what the walk parses and reference-validates, while scan.maxNodes (default 256) now caps only the graph render. References resolve across the whole corpus, so large repos no longer flag links to unrendered files as broken. Adds the --max-scan flag and the /api/folders, /api/branch, and /api/scan?meta=1 endpoints that back the lazy folders tree and branch-scoped map.

  User-facing

  Large repos now scan and validate references across the whole tree; check folders (with per-folder issue counts) to choose what the map shows. Map palettes count what is shown; a Reset filters button clears it all; the refresh button spins while any scan runs.

Patch Changes

• Restores the files rail's per-row stale-clock icon, dropped when the rail switched to building from the lightweight GET /api/folders payload (which carried the error / warn counts but not the sidecar drift status). The endpoint now emits a sidecarStatus field (the persisted scan_nodes.sidecar_status, null when there is no parseable sidecar), threaded from the kernel loader through the BFF into the rail so staleness flags corpus-wide in demo and sm serve mode.

  User-facing

  The files rail again flags out-of-date nodes with the clock icon, so you can see at a glance which files have drifted since their last review.

• Body extractors now strip raw HTML (comments and tag tokens) before matching, alongside the existing code-region strip. A markdown link commented out as <!-- [x](old.md) --> or hidden in an attribute value (<img alt="[x](y.md)">) no longer produces a phantom edge. The strip is bounded to comments and tag tokens, so markdown nested inside a <div> block still resolves; core/backtick-path is unaffected (HTML is not a code region).

  User-facing

  Scanning .md files that contain HTML no longer creates phantom links or false broken-reference warnings from links that were commented out or tucked inside HTML attributes.

@skill-map/cli@0.62.2

Patch Changes

• The /api/branch map projection now keeps an edge when its RESOLVED target is a rendered node, not only when the raw authored target is. Trigger-style invokes / mentions links store the trigger (/cmd, @agent) in target and the real node path in resolvedTarget; the old filter matched the raw target alone, so every resolved trigger edge was dropped from the graph and the map showed only path-style references. Genuinely-broken links (no resolved node) stay excluded.

  User-facing

  The graph map again draws invokes and mentions arrows (a command running a skill, an agent referenced by name), not just plain file references. A recent change had hidden every resolved trigger edge from the map.