agentleFS
Sign inStart free
SantiagoVqzSantiagoVqz/claude-skills/skills/support/ask/SKILL.md

ask

Skill2 starsChanged 11 days agoMITView on GitHub

Pick the skill of this set, or the flow, for the work at hand, from the state of the repo. Use when the user runs /ask, asks which skill fits or what to run next, or at the start of work when no skill of this set was named.

72 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use daily-arxiv, exp-eval and 10 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.
  • LambenthanSkill

    daily-arxiv

    每日从 arXiv 拉取新论文，过滤相关性，auto-ingest 高优先级论文，检测 SOTA 更新

    Also in 5 projects that use this

  • LambenthanSkill

    exp-eval

    实验判决门：Review LLM 独立评判实验结果 → 4 种判决路径 → 自动更新 claims confidence、ideas status、graph edges

    Also in 5 projects that use this

  • skyllwtSkill

    exp-pilot-eval

    Pilot result evaluation — read pilot results, apply success criteria, update idea page (pilot_result, failure_reason if failed), generate PILOT_VERDICT_REPORT. Called by /ideate Phase 5 after /exp-pilot-run.

    Also in 4 projects that use this

  • skyllwtSkill

    exp-pilot-run

    Pilot experiment execution — read Pilot Spec YAML, write pilot code, run experiment(Confirm with the user before operation and require the applicant to conduct manual inspection), return results. Called by /ideate Phase 5. Does NOT modify wiki pages or judge pass/fail.

    Also in 4 projects that use this

  • phanghonghaoSkill

    poster

    HTML/CSS A4 海报生成。从沙龙/活动 Markdown 文件生成科技风 A4 海报 PDF/PNG。Use when user mentions: poster, 海报, 海报生成, make poster, generate poster, 活动海报, 沙龙海报.

    Also in 4 projects that use this

  • +6 more from Lambenthan
  • +1 more from skyllwt

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 3 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 60 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When no branch matches, say so, and give the two nearest skills with the difference between them.

    line 64
  2. Distinctive:

    When the state skips a step of the flow, name the step and the fact that skips it.

    line 78
  3. Distinctive:

    When the user asks for a step out of order, name the command that the flow puts first and its reason.

    line 80
  4. Very specific:

    Every skill runs alone, with no earlier step.

    line 93

“Common” is measured: kept by at least 6 in 10 of the 60 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

Ask

This skill is a router over the skills of this set. It names the next command and the reason. It runs no other skill and changes nothing.

The summaries here are for orientation. Before you say what a skill does, or that a step can be skipped, read that skill's SKILL.md.

1. Read the state

Find the facts below. Use git, gh and the files. Ask the user only for a fact that the repo cannot give.

  • Branch: the current branch: fix/<n>-<slug>, the trunk, or other.
  • Git operation: a rebase or a merge stops on a conflict, or none is in progress.
  • Issue: the issue in the branch name or the request, its state and its kind. For a spec, its open tickets and their blockers.
  • PR: for a branch that is not the trunk, an open PR, a merged PR, or none.
  • Request: what the user wants, in one line.

Completion: you have each fact, or you know that it does not apply.

2. Route

Go down the tree. The first match is the answer.

A rebase or merge stops on a conflict ─────────► resolve-conflicts
File or track work for later (idea, bug) ──────► /issue
Branch has a merged PR ────────────────────────► /cleanup
Branch has an open PR ─────────────────────────► you merge, then /cleanup
Something is broken (error, Sentry, bug) ──────► /diagnose
fix/ branch with commits, no PR ───────────────► /ship
Spec issue with open tickets ──────────────────► /build <ticket with no open blocker>
Any other ticket or task, or a small change ───► /build <issue or words>
New idea or feature with stories to settle ────► /spec
Research, look up or compare a topic ─────────► /research
A design question that talk cannot settle ─────► prototype
Stress-test a plan or decision ────────────────► interview
Check a diff ──────────────────────────────────► /verify
Check a change in a browser ───────────────────► /ui-check
Change structure, keep behavior ───────────────► refactor
Build one behavior test-first, with no spec ───► tdd
Write a PR body ───────────────────────────────► pr
Make, provision or remove a worktree ──────────► worktree
Once a week of work ───────────────────────────► /retro
Edit a skill, AGENTS.md or CLAUDE.md ──────────► writing-for-agents
Cut AI tells from text ────────────────────────► unslop
Explain a ticket in plain terms ───────────────► /brief
A manual step: keys, CI secrets, a cutover ────► /wizard

When no branch matches, say so, and give the two nearest skills with the difference between them.

Completion: you have one command, or one command for each part of a request with several parts.

3. Answer

Give the answer in this shape:

Next: /<skill> <argument>
Why: <the facts from step 1 that chose it>
Then: <the command after it in the flow>

When the state skips a step of the flow, name the step and the fact that skips it.

When the user asks for a step out of order, name the command that the flow puts first and its reason. The user decides.

The flows

Feature:  /spec ──► /build, once per ticket ──► /ship ──► you merge ──► /cleanup
Ticket:   you write it ──► /build ──► /ship ──► you merge ──► /cleanup
Bug:      /diagnose ──► /ship ──► you merge ──► /cleanup
  • Each /build works in your session, on one ticket: plan, one approval, build, scoped checks. A hard stop gets the full plan in that session.
  • /build picks tdd, refactor or diagnose from the work, or builds directly.
  • /ship runs the hard-stop check, the one cold read and the full check.
  • Every skill runs alone, with no earlier step. /setup-skills is optional.
  • Skills with no slash in this file are called by other skills. The user can also run them directly.

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.

Check your setup, free