agentleFS
joris887joris887/exosuit/.claude/skills/doctor/SKILL.md

doctor

Skill9 starsChanged 3 months agoMITView on GitHub

Validate framework configuration, check runtime dependencies, and report issues. Use when something isn't working or after setup.

  • Reads credentials

85 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use cancel, hud and 9 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • woosikkimSkill

    cancel

    Clear any stale OMC-Slim state

    Also in 4 projects that use this

  • ibahgatSkill

    hud

    Control omi visual HUD profile and keep status output readable in normal, compact, or hidden mode.

    Also in 3 projects that use this

  • ibahgatSkill

    ralplan

    Create a strict, quality-first execution plan with explicit gates and rollback-ready sequencing.

    Also in 3 projects that use this

  • QinghongLinSkill

    editor

    Read analyst.json and detective.json, make all editorial decisions — what the blog argues, which findings matter, narrative arc and section structure. No visual design. Outputs editor.md (prose) and editor.json (structure with edt_xx IDs).

    Also in 3 projects that use this

  • BlockSkill

    open

    Open markdown documents in Penpal for collaborative review. Use when you want a human to review a document, plan, or research artifact.

    Also in 3 projects that use this

  • Arman-KudaibergenovSkill

    inspect

    Этот скилл MUST быть вызван для анализа и инспекции любого объекта 1С из XML-выгрузки: конфигурация (cf-info), объект метаданных (meta-info), управляемая форма (form-info), схема компоновки данных СКД (skd-info), макет табличного документа MXL (mxl-info), подсистема (subsystem-info), роль и права (role-info). SHOULD вызывать как подготовительный шаг перед любой модификацией объекта. Do NOT использовать для модификации — для изменений используй соответствующие edit/compile скиллы.

    Also in 3 projects that use this

  • +1 more from ibahgat

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. 0 are common, 6 are distinctive, and 8 are very specific to this project.

  • 0

    Common

    Most of the 64 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 8

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run a comprehensive health check on the framework configuration and project setup.

    line 16
  2. Distinctive:

    Read CLAUDE.md Commands section.

    line 20
  3. Very specific:

    Run the command in dry-run or check mode (e.g., npm test --dry-run, ruff check --quiet .).

    line 23
  4. Distinctive:

    Verify the hook script file exists and is executable

    line 34
  5. Very specific:

    Read the YAML frontmatter paths: field

    line 45
  6. Distinctive:

    Check if any files in the project match those patterns

    line 46
  7. Very specific:

    Check docs/sessions/.activity-log.jsonl for rule trigger frequency over the last 5 sessions

    line 48
  8. Very specific:

    Verify each dependency skill exists in .claude/skills/

    line 56
  9. Very specific:

    Verify each called skill exists in .claude/skills/

    line 62
  10. Distinctive:

    Check key framework files:

    line 68
  11. Very specific:

    Verify lib/hook-guard.sh exists and is executable

    line 136
  12. Very specific:

    If docs/reference/READINESS_REPORT.md exists, compare the bootstrap baseline against current project state:

    line 141
  13. Distinctive:

    Read and parse the readiness assessment table (Principle / Status / Detail)

    line 143
  14. Very specific:

    If the file does not exist, report: "No readiness baseline — run /bootstrap to generate docs/reference/READINESS_REPORT.md."

    line 160

“Common” is measured: kept by at least 6 in 10 of the 64 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

doctor

Run a comprehensive health check on the framework configuration and project setup.

1. Command Verification

Read CLAUDE.md Commands section. For each configured command (test, lint, format, build, typecheck):

<IF condition="command is configured (not placeholder)"> Run the command in dry-run or check mode (e.g., `npm test --dry-run`, `ruff check --quiet .`). Report: PASS if it executes, FAIL if it errors, SKIP if not configured. </IF> <ELSE> Report: NOT CONFIGURED — suggest running `/bootstrap` to detect. </ELSE>

2. Hook Dependencies

For each hook in .claude/settings.json:

  • Verify the hook script file exists and is executable
  • For post-edit-format.sh: check if at least one formatter is installed (prettier, biome, ruff, black, rustfmt, gofmt)
  • For shell hook scripts: verify git is available (used by stop.sh, session-start.sh)
  • Optional: check if jq is available (hooks fall back to sed without it)

Report: PASS/WARN (missing optional tools)/FAIL (missing required tools)

3. Rule Relevance

For each rule in .claude/rules/:

  • Read the YAML frontmatter paths: field
  • Check if any files in the project match those patterns
  • Report: ACTIVE (matches found) / DORMANT (no matches — rule will never trigger)
  • Check docs/sessions/.activity-log.jsonl for rule trigger frequency over the last 5 sessions
  • Rules that are ACTIVE but have zero trigger events in 5+ sessions: report as REVIEW (candidate for retirement)
  • The "pull their weight" test: if a rule never changes AI behavior, it wastes context budget

4. Skill Dependencies & Cross-References

For each skill with depends-on: in YAML frontmatter:

  • Verify each dependency skill exists in .claude/skills/
  • For skills with requires: — check binaries, commands, and files
  • Report: PASS/WARN (optional dependency missing)/FAIL (required dependency missing)

For each skill with calls: in skills-registry.json:

  • Verify each called skill exists in .claude/skills/
  • Report any broken references (skill calls a non-existent skill)
  • Report orphaned skills (not called by any other skill AND not user-invocable) as INFO — may be intentionally standalone

5. Documentation Freshness

Check key framework files:

  • CLAUDE.md: Is "Current Focus" section filled (not placeholder)?
  • docs/progress.md: Has it been updated this sprint?
  • docs/reference/BACKLOG_INDEX.md: Are status counts current?
  • docs/reference/GROUND_RULES.md: Does it exist and have principles?

Report: CURRENT/STALE/MISSING for each

6. Git State

  • Branch naming follows convention (feat/*, fix/*, hotfix/*, refactor/*, docs/*, test/*, chore/*, or sprint-*)?
  • Any stale branches (merged but not deleted)?
  • Any unmerged branches with no commits in the last 2 weeks?
    git for-each-ref --sort=-committerdate --format='%(refname:short) %(committerdate:relative)' refs/heads/ | while read branch date; do echo "$branch — last commit $date"; done
    Flag branches with no commits in >14 days as "potentially stale — consider merging or deleting"
  • Remote configured and accessible?

7. Skill Conformance

Execute scripts/validate-skills.sh directly — do NOT read source first.

bash .claude/skills/doctor/scripts/validate-skills.sh

Reports per-skill conformance: YAML frontmatter, line budget, required sections, reference budgets, registry version match. Include results in the health report under a "Skill Conformance" section.

8. Skill Usage Analytics

Parse docs/sessions/.activity-log.jsonl for entries with "type":"skill":

# Extract skill invocation data from activity log
grep '"type":"skill"' docs/sessions/.activity-log.jsonl 2>/dev/null || echo "No skill data"
  • Count invocations per skill name over all available log data
  • Find the most recent invocation timestamp per skill
  • List all skill directories from .claude/skills/ (excluding SKILLS_INVENTORY.md, SKILL_TEMPLATE.md, and non-directory entries)
  • Cross-reference invocations with available skills to classify:
    • ACTIVE: Invoked at least once in the log
    • DORMANT: Skill directory exists, user-invocable, but never invoked
    • FREQUENT: Invoked 5+ times (top skills)
  • Note: Skills not in the log may still be useful — they may not have been needed yet. DORMANT is informational, not a problem.

9. Profile & Hook Status

Report the current project profile and hook configuration:

echo "Project profile: $(cat .claude/hooks/state/project-profile 2>/dev/null || echo 'standard (default)')"
echo "EXOSUIT_PROJECT_PROFILE=${EXOSUIT_PROJECT_PROFILE:-<not set, reading from CLAUDE.md>}"
echo "EXOSUIT_HOOK_PROFILE=${EXOSUIT_HOOK_PROFILE:-<derived from project profile>}"
echo "EXOSUIT_DISABLED_HOOKS=${EXOSUIT_DISABLED_HOOKS:-<none>}"
echo "EXOSUIT_EXPLAIN_MODE=${EXOSUIT_EXPLAIN_MODE:-brief}"
echo "EXOSUIT_STOP_MAX_ITERATIONS=${EXOSUIT_STOP_MAX_ITERATIONS:-5}"
  • Show current project profile and its source (env var, CLAUDE.md **Profile:** line, or default)
  • Show derived hook profile and whether it was explicitly overridden
  • Profile implications:
    • lean: minimal hooks, simplified story-cycle (Plan→Build→Verify), no quality agents, no sprint specs
    • standard: standard hooks, full 8-phase story-cycle, 3 quality agents, full docs
    • strict: strict hooks, all phases mandatory for all stories, 5 quality agents + integration-tester, audit trail
  • List any disabled hooks
  • Verify lib/hook-guard.sh exists and is executable
  • Show effective stop iteration limit (strict defaults to 10)

10. Readiness Progress

If docs/reference/READINESS_REPORT.md exists, compare the bootstrap baseline against current project state:

  1. Read and parse the readiness assessment table (Principle / Status / Detail)
  2. For each principle classified as ⚠️ Risk or ✗ Missing, re-evaluate using data from sections 1–9:
    • TDD-first → Test command (§1) + run coverage tool if available
    • Sprint-based / Git-disciplined → Git state (§6)
    • Verification-driven → Test command passes (§1)
    • CI-enforced → Check for .github/workflows/, .gitlab-ci.yml, .circleci/, Jenkinsfile
    • Secrets-aware → Post-edit hook (§2) + .env* in .gitignore
    • Anti-slop → code-slop.md rule exists (§3)
    • Quality gates → Formatter + linter + coverage + typecheck all available (§1)
    • Context-efficient → Source files >500 LOC: find . -name '*.{ext}' -not -path '*/node_modules/*' | xargs wc -l | awk '$1>500 && !/total$/' | wc -l
    • Documentation-lean → Core docs populated (§5)
    • Pre-commit hooks → .pre-commit-config.yaml, .husky/, or lefthook.yml exists
    • Type-safe → Typecheck command works (§1)
    • Contract-first / API-documented → API spec files + API_DOCUMENTATION.md (skip if no API detected)
    • Decisions-documented → docs/adr/ has ≥1 accepted ADR
  3. Classify each as ✓ Ready, ⚠️ Risk, or ✗ Missing using bootstrap's classification rules

If the file does not exist, report: "No readiness baseline — run /bootstrap to generate docs/reference/READINESS_REPORT.md."

Output

## Framework Health Report

### Commands
| Command | Status | Notes |
|---------|--------|-------|
| test | PASS/FAIL/NOT CONFIGURED | [details] |
| lint | PASS/FAIL/NOT CONFIGURED | [details] |
| ... | | |

### Hooks
| Hook | Status | Missing Tools |
|------|--------|---------------|
| post-edit-format.sh | PASS/WARN | [list] |
| ... | | |

### Rules
| Rule | Status | Matching Files |
|------|--------|----------------|
| testing.md | ACTIVE | [count] files |
| ... | | |

### Skills
| Check | Status | Details |
|-------|--------|---------|
| Dependency resolution | PASS/WARN | [details] |
| Prerequisites | PASS/WARN | [details] |

### Documentation
| File | Status | Last Updated |
|------|--------|-------------|
| CLAUDE.md | CURRENT/STALE | [date] |
| ... | | |

### Skill Conformance
| Skill | Frontmatter | Lines | References | Registry |
|-------|-------------|-------|------------|----------|
| [name] | PASS/FAIL | [count] | PASS/WARN | PASS/WARN |

### Skill Usage
| Skill | Invocations | Last Used | Status |
|-------|-------------|-----------|--------|
| story-cycle | 15 | 2026-03-25 | ACTIVE |
| manual-test | 0 | never | DORMANT |

### Profile & Hook Configuration
| Setting | Value | Source |
|---------|-------|--------|
| Project Profile | lean/standard/strict | env var / CLAUDE.md / default |
| Hook Profile | minimal/standard/strict | env var / derived from project |
| Disabled Hooks | none or list | EXOSUIT_DISABLED_HOOKS |
| Explain Mode | off/brief/verbose | EXOSUIT_EXPLAIN_MODE |
| Stop Iterations | 5/10/custom | EXOSUIT_STOP_MAX_ITERATIONS |
| Guard script | PASS/FAIL | lib/hook-guard.sh |

### Readiness Progress (if baseline exists)
| Principle | Bootstrap | Current | Progress |
|-----------|-----------|---------|----------|
| TDD-first | ✗ Missing | ✓ Ready | ↑ Fixed |
| CI-enforced | ⚠️ Risk | ⚠️ Risk | → Unchanged |

**Progress:** X of Y gaps addressed since bootstrap ({date})
_Only rows for principles that were Risk or Missing at bootstrap. Omit Ready principles._

### Overall: X/Y checks passed — [HEALTHY / NEEDS ATTENTION / ACTION REQUIRED]

Next Steps

Based on findings, suggest:

  • /bootstrap — if commands or hooks need configuration
  • Specific fixes for any FAIL items
  • /weekly-maintenance — if documentation is stale
  • Foundation stories from E00-foundation.md — if readiness gaps persist

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.