Agent skill
repo-capability-map
Analyzes repository modules for current capabilities, groups them by domain, and identifies strategic gaps against mission objectives
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/repo-capability-map
SKILL.md
Repo Capability Map
Discover what each repository can do, classify capabilities by domain, assess maturity, and identify strategic gaps.
Purpose
While repository-health-analyzer asks "is the repo well-maintained?", this skill asks "what can this repo do, and what's missing?"
It produces a structured capability inventory across the entire workspace, enabling:
- Strategic planning against mission objectives
- Identification of capability gaps before they become blockers
- Cross-repo dependency awareness
- Maturity tracking over time
Capability Taxonomy
Two-level hierarchy: Domain > Capability Area
| Domain ID | Domain Name | Description |
|---|---|---|
ENG |
Engineering Analysis | Simulation, modeling, structural/fatigue analysis |
DATA |
Data & Analytics | Data pipelines, visualization, statistical analysis |
WEB |
Web & Frontend | Websites, UI components, client-side applications |
BIZ |
Business Operations | Admin tools, invoicing, client management |
INFRA |
Infrastructure & DevOps | CI/CD, packaging, deployment, shared utilities |
KNOW |
Knowledge & Documentation | Media, resumes, knowledge bases, training materials |
FIN |
Financial & Investments | Portfolio tracking, financial modeling, reporting |
The taxonomy is discovered dynamically -- the agent maps findings to known domains but can create new ones if a repo doesn't fit.
Maturity Model
| Level | Label | Criteria |
|---|---|---|
| 1 | Nascent | Mentioned in mission/roadmap but no implementation exists |
| 2 | Emerging | Partial implementation, <50% test coverage, active WIP |
| 3 | Established | Working implementation with tests, docs, CLI/API available |
| 4 | Mature | Production-grade, 80%+ coverage, skills defined, decision history |
Assessment is based on observable artifacts (files exist, tests present, skills defined) -- not subjective judgments.
Analysis Methodology
For each repo, read files in priority order:
Tier 1 -- Declared Identity (highest confidence)
README.md-- mission, capabilities snapshot, featuresCLAUDE.md-- tech stack, project rules, domain keywords.agent-os/product/mission.md-- strategic phases, success metrics.agent-os/product/roadmap.md-- planned features, completion status
Tier 2 -- Structural Evidence
src/directory tree -- module names, sub-packages.claude/skills/-- domain skills (each skill = capability signal)pyproject.toml/package.json-- dependencies, entry pointstests/anddocs/structure
Tier 3 -- Behavioral Evidence
git log --oneline -30-- recent activity, development focusscripts/directory -- automation capabilitiesdata//config/contents -- data domains
Actions
| Action | File | Description |
|---|---|---|
| scan | actions/scan.md |
Extract capabilities from one or all repos |
| gaps | actions/gaps.md |
Analyze gaps against mission objectives |
| report | actions/report.md |
Generate console summary + markdown report |
| compare | actions/compare.md |
Side-by-side comparison between two repos |
Templates
| Template | File | Description |
|---|---|---|
| Repo Profile | templates/repo-profile.md |
Per-repo capability profile |
| Capability Report | templates/capability-report.md |
Full workspace capability report |
Integration Points
- repository-health-analyzer -- health gates capability scanning (unhealthy repos get flagged)
- work-queue -- gap analysis can feed
/work additems for uncovered objectives - claude-reflect -- periodic reflection includes capability changes over time
- compliance-check -- repos missing CLAUDE.md receive lower maturity scores
Command Reference
See @.claude/commands/workspace-hub/repo-capability-map.md for usage.
Didn't find tool you were looking for?