agentleFS
geekxxgeekxx/ai-assistant-starter-kit/.claude/skills/intake/SKILL.md

intake

Skill1 starsChanged 4 months agoMITView on GitHub

Delegation intake interview. Use before delegating any significant or ambiguous task to a specialist. Runs a 6-question interview that forces precision, then compresses the answers into the spawn prompt the specialist receives. Skip for small, unambiguous tasks (filing, list updates, lookups).

34 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use triage, plan 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.
  • Google CloudSkill

    triage

    Gather context on a change or bug and plan the implementation.

    Also in 5 projects that use this

  • BlockSkill

    plan

    Draft a TDD, PRD, or lightweight plan in the conversation without committing it.

    Also in 4 projects that use this

  • Microsoft AzureSkill

    build

    Build the TypeSpec Rust emitter from source. Run this after cloning the repo, after pulling changes, or when TypeScript source files have been modified.

    Also in 3 projects that use this

  • AWSSkill

    retro

    Conduct a retrospective on a coding session.

    Also in 4 projects that use this

  • VercelSkill

    deploy

    Deploy applications to Vercel. Use when the user says "deploy", "deploy to Vercel", "push to production", "deploy my app", or "go live".

    Also in 3 projects that use this

  • Microsoft AzureSkill

    verify

    End-to-end app testing pipeline for browser-based web apps on cloud browsers via Azure Playwright Workspaces: verify the live app, author tests for working flows, heal failing tests, then re-run with video and produce one consolidated report. Three phases (app verification, test authoring, test healing) plus a scoped final sanity run; each phase emits a sub-report. Use when asked to "verify my app", "make sure my app works and has good test coverage", "do a full QA pass", "add tests where I'm missing coverage", "fix failing tests and add new ones", "smoke test before deploy", "QA my deployment", or any combined verify-and-test ask. Requires Azure CLI (`az login`) and a Playwright Workspace resource ID — bootstraps configs on first invocation. Browser-based web apps with Playwright only — not for unit/API tests, native mobile/desktop, load/perf, or full accessibility audits.

    Also in 4 projects that use this

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 7 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 33 edited versions we read keep these.

  • 7

    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:

    Run it when the task is significant or the brief is ambiguous: a deck, a research piece, a multi-step deliverable, anything where a wrong …

    line 14
  2. Distinctive:

    Do not: run it for small, unambiguous tasks (filing a note, updating a list, a lookup, archiving an inbox file).

    line 16
  3. Distinctive:

    When in doubt, ask Jeff: "This one's worth a quick intake.

    line 18
  4. Distinctive:

    Ask one question at a time.

    line 22
  5. Distinctive:

    After each answer, reflect back what you heard in one sentence, then continue.

    line 23
  6. Distinctive:

    Persist durable shared language to memory: Check whether any term from Q3 or decision from Q4 is worth keeping across sessions.

    line 47
  7. Distinctive:

    Before writing a memory, apply the filter:

    line 50
  8. Very specific:

    Only persist entries that pass all three.

    line 55

“Common” is measured: kept by at least 6 in 10 of the 33 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

Intake — Delegation Brief

A specialist you spawn sees only the prompt you hand them. No session history, no memory of past conversations, no context except what you write. This interview turns that delegation brief from a guess into a decision.

When to run this

  • Run it when the task is significant or the brief is ambiguous: a deck, a research piece, a multi-step deliverable, anything where a wrong interpretation costs real rework.
  • Do not run it for small, unambiguous tasks (filing a note, updating a list, a lookup, archiving an inbox file). The friction would just train Jeff to skip it.
  • When in doubt, ask Jeff: "This one's worth a quick intake. Two minutes?" then proceed.

The interview

Ask one question at a time. Wait for the answer before moving on. After each answer, reflect back what you heard in one sentence, then continue. The harder a question is to answer, the more useful it is — that difficulty is the language gap a specialist would otherwise fill with a guess.

  1. The task — What are you trying to make, decide, or get done? One sentence. If it takes more than one sentence, the scope isn't clear yet.
  2. Done looks like — What's the deliverable that lands in owners_inbox/, in what format, and what is the checkable exit condition: the specific test, rubric, count, or sign-off that tells the specialist to stop iterating? A vague aspiration ("it should feel right") is not an exit condition — push for something that can pass or fail. (Coaching: if Jeff's answer is fuzzy, ask: "How would you decide in under 60 seconds whether the specialist's draft is accepted or needs another pass?" Press until the answer is testable.)
  3. Shared language — What terms, named things, people, or prior decisions does this touch that the specialist wouldn't know cold? For each: what does it mean in your context?
  4. Locked decisions — What's already decided and not up for re-litigation? For each: the decision, and why it's locked.
  5. The failure mode — What does a bad deliverable look like? What's the real cost of getting it wrong? This becomes the guardrail in the spawn prompt.
  6. The one thing — If this has to get one thing right above all else, what is it? This becomes the headline instruction to the specialist.

After the interview

  1. Route. Name the specialist who should own this (or flag that a new hire is needed via Scout + Forge). Stay the single point of contact.
  2. Compress into the spawn prompt. Fold the six answers into the delegation prompt:
    • Q1 → the objective line
    • Q2 → deliverable spec + explicit exit condition (the checkable test the specialist uses to know when they are done) + output path
    • Q3 → a "context the specialist won't have" section
    • Q4 → constraints, marked as locked
    • Q5 → explicit guardrails / "do not" list
    • Q6 → the single headline instruction, stated first
    • Pick the model per CLAUDE.md (task complexity vs. cost).
  3. Persist durable shared language to memory. Check whether any term from Q3 or decision from Q4 is worth keeping across sessions. Use your assistant's existing memory system — do not create a separate context file. Before writing a memory, apply the filter:

    • Would you have to explain this again next session without it?
    • Is it genuinely recurring, not specific to this one task?
    • Does it already exist in memory or CLAUDE.md?

    Only persist entries that pass all three. A shared-language term or locked decision is usually type: reference or type: project; keep it tight (name, one-line definition, and Why: for a decision). When in doubt, don't write it.

Then delegate, using the interview answers as the specialist's context.

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.