Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/haruki1090-doctor
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/doctor
What's inside
- Skill name
doctor- Description
- Check plumb installation and routing health. Use when asked to "run doctor", "check plumb's health" or "is the harness broken", or after relevant environment components change.
- License
- MIT
- Instructions
- 52 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# doctor
```bash
plumb-doctor
```
**The script holds the judgment. This document holds only when to run it, and what to suspect when it fails.**
Copy the steps here and one copy goes stale (**principle-encode-lessons-in-structure**).
## Why it is needed
plumb is **a bundle of documents that assert facts about the environment**:
which command exists, which skill exists, which path exists.
**Unlike code, those assertions say nothing when they break.**
A document happily survives the disappearance of what it points at, and sends the next reader somewhere that is not there.
Two of them turned up in a single session on 2026-08-29.
```
~/.<tool>-worktrees/ a carry-over note named it; the path did not exist
the pr-review skill six agents introduced themselves as "called from stage N of the
pr-review skill", but that skill had been moved aside and was not live
```
**Neither shows up under grep or lint.** "Something that should be there is not"
is only knowable by going and checking that it is. So it became a tool (**principle-build-the-lever**).
## When to run it
- **At the end of a batch.** As part of the termination condition.
- **After you change the environment.** A skill added, removed or moved aside; an extension installed; a CLI swapped.
- **When you come back to plumb after a while.** When you do not know what moved since last time.
- **When you followed a pointer out of the `plumb` index and it was not there.** Do not fix that one item; look at the whole first.
## Reading a failure
| Where it failed | What to suspect |
|---|---|
| Inside (cross-cutting rules) | an edit to plumb itself. Run `plumb-check` directly and read what it says |
| Inside (script behavior) | selftest.sh's own premises may be broken. Run `plumb-selftest` directly and read what it says |
| Routing targets | PATH and the extension install. The premises of `docs/role-map.md` have collapsed |
| Bundled agents | one of `agents/` was deleted or renamed. Cross-check the names `skills/pr-review/SKILL.md` calls against what is actually in `agents/` |
| The agents on your side (your own copy) | the same cause, showing up on your own `~/.claude/agents/` side. Either put the skill back, or fix the agent's description |
| Paths | `docs/path-map.md`'s claim is out of date. **Measure first, then fix the table. Do not rewrite it from a guess** |
| Loading | check the plugin list for the selected runtime in a new session; use the runtime adapter for command mapping |
**Do not bury a failing item.** Loosening the expectation to get doctor to pass is
killing the verifier (**principle-prove-it-works**).
If you cannot fix it, write "this is broken right now" into `docs/` and leave it there.From Haruki1090/plumb, MIT.
More from Haruki1090/plumb
- 0.570score
plumb
Haruki1090
Route complex engineering work to plumb playbooks and principles.
SkillAFreshness: AClaude Code +9 - 0.570score
graph
Haruki1090
Design dependencies, shared resources, and verification for complex multi-step engineering work; use when an execution graph or parallel plan is needed.
SkillAFreshness: AClaude Code +9 - 0.570score
lead
Haruki1090
Coordinates a queue of pull requests or issues end to end.
SkillAFreshness: AClaude Code +9 - 0.570score
pr-review
Haruki1090
Review a pull request as the approver, checking its claims against the diff with independent refutation and reproduction.
SkillAFreshness: AClaude Code +9 - 0.567score
setup
Haruki1090
Codex only: not for Claude Code.
SkillAFreshness: AClaude Code +9 - 0.561score
interrogate
Haruki1090
Add a different model family to the refutation stage of a PR review when that independent perspective is requested or warranted.
SkillAFreshness: AClaude Code +9 - 0.559score
decision-brief
Haruki1090
Compare meaningful alternatives when a user decision changes what will be built, using concrete evidence and clear tradeoffs.
SkillAFreshness: AClaude Code +9 - 0.556score
plumb-codex
Haruki1090
Codex only: not for Claude Code.
SkillAFreshness: AClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Sep 28, 2026 (1 month ago)
- Last release
- Oct 2, 2026 (27 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/Haruki1090/plumb
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from Haruki1090
- 0.570score
plumb
Haruki1090
Route complex engineering work to plumb playbooks and principles.
SkillAFreshness: AClaude Code +9 - 0.570score
graph
Haruki1090
Design dependencies, shared resources, and verification for complex multi-step engineering work; use when an execution graph or parallel plan is needed.
SkillAFreshness: AClaude Code +9 - 0.570score
lead
Haruki1090
Coordinates a queue of pull requests or issues end to end.
SkillAFreshness: AClaude Code +9 - 0.570score
pr-review
Haruki1090
Review a pull request as the approver, checking its claims against the diff with independent refutation and reproduction.
SkillAFreshness: AClaude Code +9 - 0.567score
setup
Haruki1090
Codex only: not for Claude Code.
SkillAFreshness: AClaude Code +9 - 0.561score
interrogate
Haruki1090
Add a different model family to the refutation stage of a PR review when that independent perspective is requested or warranted.
SkillAFreshness: AClaude Code +9
Cite this page
doctor skill. Agent Mart. https://agentmart.dev/skills/haruki1090-doctor (last updated 2026-10-24)