agentleFS
Sign inStart free
purpleeddypurpleeddy/cladding/skills/doctor/SKILL.md

doctor

Skill16 starsChanged 4 months agoMITView on GitHub

Diagnose Cladding runtime health — Claude Code hook liveness and version, CI package pinning, lifecycle governance, and sentinel-miss frequency by phase × cause × fallback. Use when hooks may be silent, CI may float across Cladding releases, scan or run results look thinner than expected, or before tuning the host model or transport. Activate only when the connected project contains spec.yaml or the user explicitly names Cladding; ignore ordinary requests in uninitialized projects.

  • Reads credentials

We split this skill into the rules it gives an agent

6 rules. 0 are common, 2 are distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 56 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run clad doctor from the project root.

    line 7
  2. Distinctive:

    When sentinel-miss events exist:

    line 19
  3. Very specific:

    After clad init --scan to confirm the scan refinement ran with full LLM coverage (no sentinel_miss events).

    line 32
  4. Distinctive:

    After installing or updating the Claude Code plugin to confirm a new session actually fired the shipped hooks and loaded the current engine.

    line 33
  5. Very specific:

    After a clad serve session to confirm the host returned refined replies through the configured dispatcher.

    line 34
  6. Very specific:

    Before reporting "the LLM seems off" to a host (Claude Code / Cursor / Continue) — the breakdown tells you whether the issue is dispatcher …

    line 36

“Common” is measured: kept by at least 6 in 10 of the 56 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

Cladding doctor

Run clad doctor from the project root. The verb is observability — it never mutates the working tree.

  • --cwd <path> — read events from a project directory other than the current one (default cwd).
  • --json — emit the raw DoctorReport shape instead of the formatted text surface; the additive shape ({cwd, events, sentinelMiss, governance, hooks, ciVersion}) is the stable wire format for MCP clients and follow-up tooling.

The text surface prints:

  1. One pulse line with total events and total sentinel-miss count (pass when zero misses, note otherwise).
  2. An event-type breakdown line (one <type>=<count> token per non-zero EventType).
  3. Claude Code hook health: whether the runtime has actually been observed, whether the observed engine version matches the current CLI, and the last firing time (or never observed) for session start, prompt submit, before edit, after edit, and session stop.
  4. A non-blocking CI warning naming each GitHub Actions workflow that invokes an unversioned or floating npx cladding package. Numeric selectors such as cladding@0.9 and cladding@0.9.4 stay quiet.
  5. Governance counts for gate runs, done attempts and rejections, stop blocks, known-failing Stop exits, blocked fingerprints reproduced by a later gate, and attestation state.
  6. When sentinel-miss events exist:
    • by phase / by cause / by fallback aggregates from the v0.3.39 telemetry payload.
    • Top-5 missed sentinels (CONVENTIONS_MD / ARCHITECTURE_YAML / SCENARIO_FLOWS / CAPABILITIES_YAML / WHY / WHAT / PURPOSE) sorted by count desc, name asc.
    • Last 3 unique dispatcher error strings (most recent first; errors are truncated to 200 chars at the emit site).
    • A one-line tuning hint.

Exit codes

  • 0 — events.log was either missing (greenfield) or readable. A greenfield workspace prints a friendly note and exits 0; a healthy host with zero misses also exits 0 with a pass line.
  • 1 — events.log.jsonl exists but cannot be parsed as JSONL (corrupt telemetry).

When to run

  • After clad init --scan to confirm the scan refinement ran with full LLM coverage (no sentinel_miss events).
  • After installing or updating the Claude Code plugin to confirm a new session actually fired the shipped hooks and loaded the current engine.
  • After a clad serve session to confirm the host returned refined replies through the configured dispatcher.
  • Periodically in CI to track miss rate across sampling-policy changes.
  • Before reporting "the LLM seems off" to a host (Claude Code / Cursor / Continue) — the breakdown tells you whether the issue is dispatcher transport (cause: dispatcher_error) or model output quality (cause: blank_section).
clad doctor
clad doctor --cwd /path/to/project
clad doctor --json

Configured-no-LLM runs (no MCP host, no ANTHROPIC_API_KEY, or --no-llm flag on clad init) do not emit sentinel_miss — those are deliberate offline runs, not misses. A doctor pass with zero events on a workspace that never reached the LLM path is therefore expected, not a problem.

Download now (ZIP)

  • Other versions75 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 75

    1. Yeachan-Heo/oh-my-codex★ 34ksame name
    2. jihe520/MathModelAgent★ 6.2ksame name
    3. spotify/portal-ai-plugins★ 2.8ksame name
    4. fakechris/obsidian_vault_pipeline★ 164same name
    5. NoobyGains/godmode★ 115same name
    6. Rich627/whatsapp-claude-plugin★ 114same name
    7. bluzir/claude-code-design★ 102same name
    8. jjanczur/tyran★ 81same name
    9. svy04/ballast★ 73same name
    10. Hainrixz/agente-pagokit★ 68same name

    65 more not listed

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • Used alongsideProjects with it also use uninstall, mode and 7 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.
    • AnthropicSkill

      uninstall

      Uninstall a community skill that was installed via the hub. Confirms before deleting files, refuses to touch first-party plugin skills, and logs every action. Use when the user wants to fully remove a community skill ("uninstall [skill]", "remove this skill") rather than just disable it.

      Also in 6 projects that use this

    • ibahgatSkill

      mode

      Inspect or switch omi operating mode/profile for speed, depth, autonomy, or strict quality gates.

      Also in 3 projects that use this

    • xiongxianfeiSkill

      editor

      Editor and translator for writing and speech with personalized linguistic learning notes. Use this skill whenever users ask to polish, proofread, translate, correct a transcript, or explain edits ("fix this", "make this sound better").

      Also in 3 projects that use this

    • VercelSkill

      setup

      Set up Vercel CLI and project configuration. Use when the user says "set up Vercel", "configure Vercel", "link to Vercel", or "vercel init".

      Also in 23 projects that use this

    • AnthropicSkill

      wrap-up

      Close out (or revisit) a Claude Managed Agent build — refresh the overview page, recap every primitive the founder now owns, show the run log and live status, suggest 1–2 tailored next upgrades, and sweep hygiene (sessions archived, key only in .env, no literal dates in deployment kickoffs). Use when the founder says "/wrap-up", "wrap up", "close it out", "where do things stand with my agent", or at the end of a /launch-your-agent build.

      Also in 4 projects that use this

    • HK-hubSkill

      distill

      Strip designs to their essence by removing unnecessary complexity. Great design is simple, powerful, and clean. Use when the user asks to simplify, declutter, reduce noise, remove elements, or make a UI cleaner and more focused.

      Also in 3 projects that use this

  • In this repository17 other files it gives its agents.

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.

Check your setup, free