triage
Skill14 starsChanged 8 days agoMITView on GitHub
Give open issues an explicit disposition (fix, close, or defer with a reason), verifying each claim and grilling where needed, and write agent briefs for the fixes. Works on one issue, one repo, or every philoserf repo.
189 other projects use a version of this.
Adopted by
Most recently taken up by
- bmtrnavsky/disciplined-vibe-codingtheir own version, first seen Nov 15, 2026
- oxc-project/oxctheir own version, first seen Nov 15, 2026
- geoffbremner/skillstheir own version, first seen Nov 15, 2026
- NickolasBenakis/agent-playbooktheir own version, first seen Nov 15, 2026
- quickchatai/quickreviewtheir own version, first seen Nov 15, 2026
- HiteSit/real-scientist-skillstheir own version, first seen Nov 15, 2026
Used with
Projects with it also use writing-beats, setup-matt-pocock-skills 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.
LesaniSkillwriting-beats
Shape an article as a journey of beats, choose-your-own-adventure style. The user picks a starting beat from the raw material, you write only that beat, then offer options for where to pivot next, beat by beat, until the article reaches a natural end. Use when the user has raw material and wants to assemble it as a narrative rather than an argument.
Also in 66 projects that use this
GoogleSkillsetup-matt-pocock-skills
Configure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.
Also in 167 projects that use this
liufeng1976Skillsetup-ts-deep-modules
Wire dependency-cruiser into a TypeScript repo so each package is a deep module — implementation hidden in subfolders, reachable only through its entry-point files. User-invoked.
Also in 42 projects that use this
SethGammonSkillascii-diagram
Generate perfectly aligned ASCII diagrams — architecture, flow, sequence, box-and-arrow. Uses a programmatic character-grid approach so alignment is guaranteed by math, not token prediction. Includes post-render verification.
Also in 4 projects that use this
simo97Skillmigrate-to-shoehorn
Migrate test files from `as` type assertions to @total-typescript/shoehorn. Use when user mentions shoehorn, wants to replace `as` in tests, or needs partial test data.
Also in 68 projects that use this
LesaniSkillwriting-shape
Take a markdown file of raw material and shape it into an article through a conversational session — drafting candidate openings, growing the piece paragraph by paragraph, arguing about format (lists, tables, callouts, quotes) at each step. Use when the user has a pile of notes, fragments, or a rough draft and wants help turning it into something publishable.
Also in 63 projects that use this
- +1 more from Lesani
Other versions189 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 189
- mattpocock/skills★ 325ksame name
- withastro/astro★ 64ksame name
- vectorize-io/hindsight★ 61ksame name
- ogulcancelik/herdr★ 47ksame name
- herdrdev/herdr★ 47ksame name
- colinhacks/zod★ 44ksame name
- QuivrHQ/quivr★ 40ksame name
- The-Vibe-Company/quivr★ 40ksame name
- vercel-labs/skills★ 39ksame name
- vercel-labs/add-skill★ 39ksame name
179 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.
In this repository31 other files it gives its agents.
- .claude/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- cc-release-reviewSkill
- verifySkill
- assisted-logSkill
- codebase-designSkill
- code-theorySkill
- code-walkthroughSkill
- diagnosing-bugsSkill
- domain-modelingSkill
- downloads-triageSkill
- editorSkill
- grillingSkill
- grill-meSkill
- grill-with-docsSkill
- implementSkill
- improve-codebase-architectureSkill
- prototypeSkill
- prSkill
- release-gateSkill
- release-shipSkill
- release-sweepSkill
- retroSkill
- rexSkill
- spec-reviewSkill
- tddSkill
- to-specSkill
- to-ticketsSkill
- which-skillSkill
- wizardSkill
- writing-for-agentsSkill
The file
We split this skill into the rules it gives an agent
7 rules. 0 are common, 7 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 177 edited versions we read keep these.
7
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 9
Every open issue in a triage pass leaves it with exactly one disposition:
- Distinctive: line 12
Close: not worth doing, already done, or a duplicate.
- Distinctive: line 35
Gather context: Read the body, comments, labels, assignees and blockers.
- Distinctive: line 37
Check for prior art: Search the code for an existing implementation of the request, by domain concept rather than the issue's wording, and …
- Distinctive: line 48
Grill if needed: If a fix isn't specified well enough to brief, call the Skill tool twice, for "grilling" and "domain-modeling", and work …
- Distinctive: line 88
Every comment triage posts starts with this marker, which is also how the next pass knows an issue was triaged:
- Distinctive: line 97
When the reporter isn't the user and the issue needs more from them, defer it with this body:
“Common” is measured: kept by at least 6 in 10 of the 177 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.
Files
Triage
Every open issue in a triage pass leaves it with exactly one disposition:
- Fix: worth doing, specified well enough to start. Post an agent brief.
- Close: not worth doing, already done, or a duplicate.
- Defer: worth keeping, not now. Post the reason and what would change it.
Issue-tracker operations and labels are in ~/.claude/references/github-tracker.md. Read it
first. Readiness is not a label: an open issue with no open blockers and no assignee is ready
to take, and /implement picks it up from there. Triage keeps that true by blocking or
assigning everything that is not ready.
Scope
The user names the scope when invoking /triage:
- One issue (
#42, or a URL): triage it. - A repo (run inside a clone, nothing named): every open issue in it.
- The workspace (run in
~/source/philoserf): every open issue in every repo one level down, skippingobsidian-vault-changelog.
A request like "what needs my attention?" means the untriaged issues in scope, oldest first. An issue is triaged when it has a triage comment (see Comments) newer than its last edit and its last non-triage comment.
Triage one issue
- Gather context. Read the body, comments, labels, assignees and blockers. Read any earlier
triage comment, and don't re-ask what it settled. Explore the code in the area, using
GLOSSARY.mdvocabulary (at the root or indocs/) and respecting ADRs where they exist. - Check for prior art. Search the code for an existing implementation of the request, by
domain concept rather than the issue's wording, and say where you looked. Search closed
issues (
gh issue list --state closed --search "<concept>") for a duplicate or an earlierwontfix, and surface any hit. - Verify the claim. For a bug, reproduce it. For an audit issue (a
**Severity:**line and a## Suggested fix), verify the Description and Location against current code: line numbers drift. The report is the specification; the suggested fix is one option among several, so design the fix yourself. Report what happened: confirmed (with the code path), refuted, or not enough detail. - Recommend. Give the disposition, the category label, and the severity where it applies, with your reasons. Wait for the user's direction.
- Grill if needed. If a fix isn't specified well enough to brief, call the Skill tool twice, for "grilling" and "domain-modeling", and work it into shape.
- Apply (see below).
Triage a pass
For a repo or the workspace, do steps 1–4 for every open issue, delegating per-repo exploration to subagents when the pass is large. Then present one table: repo, issue, title, disposition, category, severity, and a one-line reason. The user confirms or overrides rows in one go; then apply them all.
An issue that /to-spec or /to-tickets produced is already specified: its disposition is
fix, and it needs no brief. Check only that its blockers are still right.
End the report with an explicit Parked list: every issue deferred, with its reason and what would unblock it.
Apply the disposition
Writes go to public issues, so make them only after the user has confirmed the disposition.
Labels follow the tracker reference: exactly one category label, a severity: label
mirroring any **Severity:** line, and only labels the repo already has.
- Fix: post an agent brief (AGENT-BRIEF.md). If the work needs a human (judgment calls, credentials, external access, manual testing), assign the user and say why in the brief; otherwise leave it unassigned.
- Close:
- Already implemented: say where it lives, then
gh issue close <n> --reason completed. - Duplicate: name the issue it duplicates, label
duplicate, thengh issue close <n> --reason "not planned". - Rejected: give the durable reason (scope, design, a trade-off), not "not now", label
wontfix, thengh issue close <n> --reason "not planned". The closed issue is the record the next prior-art search finds.
- Already implemented: say where it lives, then
- Defer: post the reason and the trigger that would bring it back. If the trigger is another issue, add a native blocking edge on it. Otherwise (a date, an upstream release, an answer from the reporter) assign the user, so the issue drops off the ready list.
Comments
Every comment triage posts starts with this marker, which is also how the next pass knows an issue was triaged:
> _This was generated by AI during triage._
## Triage: <fix | close | defer>When the reporter isn't the user and the issue needs more from them, defer it with this body:
**What we've established so far:**
- point 1
**What we still need from you (@<reporter-login>):**
- question 1Questions must be specific and answerable. Once the reporter replies, the issue counts as untriaged again.
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
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.

