Skills

Skills

Discover and search installed global skills plus enabled project-local pack skills, grouped by workflow stage or activity type.

Process

1. Parse arguments:

   • If $ARGUMENTS is empty or list → list mode (show all skills grouped by workflow stage).
   • If $ARGUMENTS is types → type mode (show all skills grouped by activity type).
   • If $ARGUMENTS starts with search → search mode (filter by keyword after search).

2. Discover skills:

   • Always scan global Codex skills in global/codex/*/SKILL.md.
   • If .agents/project.json exists, read enabled_packs and scan packs/<pack>/codex/*/SKILL.md for each enabled pack.
   • Also scan all packs in packs/*/codex/*/SKILL.md to discover available-but-not-installed skills.
   • If .codex/skills/*/SKILL.md exists in the current project, include those local skills too.
   • Read the first 6 lines of each file to extract YAML frontmatter fields: name, description, type.
   • Track which pack each skill belongs to (if any).

3. Group skills by workflow stage using this static mapping:

   The mapping includes global skills and pack-provided skills. Only skills discovered in step 2 should be printed, so domain and kanban entries appear only when the corresponding project-local pack is enabled or linked.

   Stage	Skills
   Pack Management	pack, pack list, pack status, pack recommend, pack install <pack-or-skill>, pack remove <pack-or-skill>, pack which <skill>
   Discovery & Market Fit	icp, enterprise-icp
   Game Development	game-workflow, game-audience, game-fantasy, game-genre-map, game-comparables, game-core-loop, game-prototype-test, game-store-page-test, game-playtest-metrics, game-roadmap, game-launch
   Devtool Development	devtool-workflow, devtool-user-map, devtool-integration-map, devtool-dx-journey, devtool-adoption, devtool-positioning, devtool-monetization, devtool-docs-audit
   Planning	brainstorm, brainstorm-kanban, idea-scope-brief, feature-interview, spec-interview, spec-interview --ideas, ui-interview, ux-variations, consolidate-variations, spec-interview-kanban, experiment, clone-spec-store (project-fleet pack)
   Mapping	journey-map, metrics
   Strategize	roadmap, roadmap-kanban, competitive-analysis, platform-strategy, gtm, landing-copy, monetization, positioning, runway-model
   Evaluate	dogfood, mvp-gap, scale-audit, customer-feedback, assumption-tracker, cohort-review, retro
   Research Health	research-roadmap, reconcile-research, reconcile-dev-docs
   Detail	plan-phase
   Execution	exec, exec-kanban
   Shipping	ship, ship-end, ship-kanban, ship-end-kanban
   Code Quality	expert-review, regression-check, dead-code
   Debugging	investigate, debug, trace, session-triage
   Refactoring & Migration	migrate, decommission, scaffold
   Monorepo	affected
   Release & Deploy	release, deploy
   Context & Session	codebase-status, analyze-sessions, handoff, sync, investor-update
   Git Workflow	branch-lifecycle
   Utility	commit-and-push-by-feature, init-agentic-skills, targeted-skill-builder, provision-agentic-config, poketo-kanban, poketo-kanban --archive, sync-roadmap-kanban, skills, risk-register

   Skills not found in the mapping go into an Other group at the end.

4. Group skills by activity type (type mode only) using the type field from frontmatter:

   Type	Description
   research	Web search + analysis → produces documents
   analysis	Reads codebase/docs → produces assessments
   planning	Interactive → produces specs/plans
   execution	Writes/modifies code
   review	Reads code → reports issues (no changes)
   debugging	Investigates + fixes problems
   shipping	Commits, deploys, session management
   ops	Manages boards, branches, meta-tooling

   Display order: research → analysis → planning → execution → review → debugging → shipping → ops. Skills with no type field go into an Other group at the end.

5. Apply filter (search mode only):

   • Match the keyword case-insensitively against each skill's name, description, and type.
   • Keep only matching skills. Remove empty groups.

6. Output results:

   • Print each non-empty group as a ## Group Name heading.
   • Under each heading, list installed/global skills as $<name> — <description>.
   • For available-but-not-installed pack skills, list as $<name> — <description> [type] ⚠ requires \$pack install ` or `$pack install ``.
   • In type mode, append the type tag: $<name> — <description> [type].
   • In list mode (stage grouping), append the type tag: $<name> — <description> [type].
   • Omit groups that have no skills (after filtering in search mode).
   • At the bottom, print a total count: **N skills** installed, **M skills** available via packs (or **N skills** matching "<keyword>").

Output Format

List mode (default — grouped by workflow stage)

## Discovery & Market Fit
$icp — Customer discovery interview...  [research]
$enterprise-icp — Enterprise multi-stakeholder discovery...  [research]

## Planning
$brainstorm — Evaluate the codebase and suggest ideas...  [planning]
...

## Planning (when a kanban pack is enabled)
$brainstorm-kanban — Brainstorm ideas and create kanban cards...  [planning]
...

**N skills** found

Type mode ($skills types)

## Research
$icp — Customer discovery interview...
$competitive-analysis — Research competitors...
...

## Analysis
$mvp-gap — Evaluate codebase against ICP...
...

**N skills** found

Constraints

• This is a read-only scan — never modify any files.
• Always read the live frontmatter from SKILL.md files; do not hardcode descriptions.
• Use the static stage mapping for stage grouping; use the type frontmatter field for type grouping.
• Keep output compact — one line per skill, no extra detail beyond name, description, and type tag.
• In search mode, if no skills match the keyword, print: No skills matching "<keyword>".

Default Shipping Contract

Follow the shared shipping contract convention in CLAUDE.md.