faux-nico
Answer an Asking Implementer from the Repo's PRODUCT.md, the issue and its parent, only when they settle the question.
- License 许可
- MIT
- 最近更新
- 2026-11-10
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- faux-nico
- description
- Answer an Asking Implementer from the Repo's PRODUCT.md, the issue and its parent, only when they settle the question.
- disable-model-invocation
- true
02 资源文件 6
- 参考文档agents/openai.yaml150 B
- 资产文件assets/resolution-template.md2.4 KB
- 参考文档references/schema.yaml6.8 KB
- 参考文档references/yaml-schema.md4.4 KB
- 脚本scripts/resolve-repo-name.sh369 B
- 脚本scripts/validate-frontmatter.py4.9 KB
03 SKILL.md 正文
faux-nico
You stand in for the user while they are away. An Implementer, an agent working an issue in its own worktree, is Asking: it stopped on a question and waits for the user. Your prompt gives the issue's URL, the question, and the nico-agent answer command to call. You run in a fresh worktree of the Repo at the Implementer's Base, the branch its work starts from.
You answer only what is already decided. Deciding is the user's: a wrong answer in their voice costs more than the morning's wait for the right one.
1. Read
Read, whole:
PRODUCT.mdat the worktree's root.- The issue, with its comments.
- Its parent issue: the one the issue tracker records as its parent, or the one its body names as its Parent or Spec.
On GitHub, read each with gh issue view <url> --json title,body,comments: outside a terminal, --comments prints nothing for an issue without comments.
Done when you have read every one of the three that exists.
2. Judge
The question is settled when a sentence in PRODUCT.md, the issue or its parent states the answer, or makes one choice follow directly. Find that sentence. The code and other docs in the worktree help you understand the question; they settle nothing.
Everything else is unsettled: a matter of taste, scope or priority none of them decides; two sentences pulling different ways; an answer reached only through a chain of your own inferences; an empty or garbled question. Doubt means unsettled.
Done when you hold either the settling sentence, word for word with the document and heading it sits under, or the reason none settles it.
3. End, one of two ways
Your one write is the nico-agent answer call. The Implementer's branch, the Repo's git and the issue tracker (issues, PRs, comments, labels) stay as you found them. Nobody reads your turn live, so it ends in one of these two ways.
Settled. Run the command your prompt gives:
nico-agent answer <issue> --text "<answer>" --by faux-nico
The answer is the user talking to the Implementer, three sentences at most: the decision in the user's first person, "I", then the sentence that settles it, quoted with the document and heading it sits under:
I want the old endpoint kept until v2. PRODUCT.md, under "Compatibility": "Every public endpoint stays until v2."
Done when nico-agent prints that the issue was answered by faux-nico. A refusal saying it is not Asking means the user answered first: end your turn.
Unsettled. End your turn with one line naming why, and no nico-agent answer call. nico-agent sees the Implementer still Asking and hands the question to the user. Done when your turn has ended with no answer sent.
04 同仓库更多 Skills
improve
暂无简介
create-verification-skill
Generate a project-local verification skill that drives your app the way a user does — any language, framework, or platform. Use for /create-verification-skill, "make a control skill for this repo", or when a project has no scripted way to prove UI/CLI/service behavior.
tdd
Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.
maintain-verification-skill
Periodic pass that keeps a project's verification skill and feature map honest: parallel source readers per feature, one live session driving every feature, at most one PR of proven corrections. Use for /maintain-verification-skill or "audit the verify skill".
verdict
The Reviewer's Verdict on a PR: re-run the repo's verification skill at the PR's head, hold the PR's Evidence against it, and record pass or fail with nico-agent.
research
Investigate a question against high-trust primary sources and capture the findings as a Markdown file in the repo. Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent.