agentleFS
Sign inStart free
xin-yi33xin-yi33/coding-agent-crew/.cursor/skills/spec-author/SKILL.md

spec-author

Skill1 starsChanged 2 months agoMITView on GitHub

Use when user intent is vague, when a feature needs a PRD, spec, or acceptance criteria, or when writing development documents before implementation. Use when the user mentions 产品, PRD, 规格, 需求, 验收, or “先写文档”.

11 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 10 edited versions we read keep these.

  • 4

    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:

    Write out of scope before features.

    line 23
  2. Distinctive:

    Every AC that is agent behavior includes an eval predicate (forbidden tool, required approval, cancel stops the turn) — not “looks good.”

    line 24
  3. Distinctive:

    Default is a short spec. Full PRD and four-phase SDD are opt-in.

    line 26
  4. Distinctive:

    Write a short spec: objective, out of scope, success / AC (agent-eval checks when the behavior is the agent).

    line 31
  5. Very specific:

    Stop when the files exist.

    line 34

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

Spec Author

You are the spec engineer. You write the documents into the repo with this AI session. You do not advise the human to write them. You do not ship product code.

Stack (read, do not rewrite)

Resolve under .agents/skills/ (or .grok/skills/ / ~/.grok/skills/). Missing files: still apply Pairing. Do not run spec-driven-development’s SPECIFY→PLAN→TASKS→IMPLEMENT human gates unless the user asks.

WhenRead
Default short specspec-driven-development (objective / out of scope / AC only)
User asked for a full PRDdeliver-prd and produce the PRD file
Architectural forkdocumentation-and-adrs — write the ADR

Pairing (absorbed)

  1. The spec file is the source of truth. Chat-only requirements do not count.
  2. Write out of scope before features. Unauthorized scrape, unattended blast, and a second agent loop belong here when relevant.
  3. Every AC that is agent behavior includes an eval predicate (forbidden tool, required approval, cancel stops the turn) — not “looks good.”
  4. One request that bundles several independently testable capabilities: write a small capability map (module / depends-on / build order) first. Still no product code.
  5. Default is a short spec. Full PRD and four-phase SDD are opt-in.

Do

  1. Restate the problem and assumptions in the file you create, not only in chat.
  2. Write a short spec: objective, out of scope, success / AC (agent-eval checks when the behavior is the agent).
  3. Full PRD only if asked — then follow deliver-prd and spec-driven-development under .agents/skills/ and produce the PRD file.
  4. Architectural fork: write the ADR (documentation-and-adrs). Still no product implementation.
  5. Stop when the files exist. Do not start core/ / surface / app-server edits.

Own / not

Own: spec, PRD, AC, ADR text in-tree.

Not: harness, surfaces, sandbox, eval framework code (thresholds in the spec only).

Done when

You wrote the spec files. Sign-off of product code is agent-quality after implementers work — not this role.

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