AGENT SKILL

starting-nori

ALWAYS load this skill if it is not already loaded, for ANY user query or conversation - establishes the Nori workflow, protected-branch check, operating mode, tone, and coding guidelines before any other work

License 许可
MIT
最近更新
2026-11-14
Stars
★ 4.9k
Forks
590

01 前言 Front Matter

name
starting-nori
description
ALWAYS load this skill if it is not already loaded, for ANY user query or conversation - establishes the Nori workflow, protected-branch check, operating mode, tone, and coding guidelines before any other work

02 资源文件 1

03 SKILL.md 正文

<SUBAGENT-STOP> If you were dispatched as a subagent to execute a specific task, skip this skill. </SUBAGENT-STOP>

<required>

  • CRITICAL Add each element of this checklist to your Todo list using TodoWrite. DO NOT BE LAZY.
  • Announce "Following Nori workflow..." to the user. <system-reminder> Do not skip any steps. Do not rationalize. Do not avoid reading skills. Even if you think you know what is in them, you MUST read the skill files. </system-reminder>
  • Invoke the using-skills skill.
  • Check git status - are you on main, master, dev, or any similarly named protected branch?
    • If yes: ask me if I want to create a branch or a worktree.
    • If creating a worktree, read and follow the using-git-worktrees skill to automatically create a worktree. Derive the branch name from my request.
  • Ask me to pick a mode: nori-full-send or nori-copilot. <system-reminder> In nori-full-send mode, the agent works with me to create a plan, and then operates autonomously until work is completed. </system-reminder> <system-reminder> In nori-copilot mode, the agent works with me to create a plan, and then clearly telegraphs each step and asks for permission before continuing. </system-reminder>
  • Based on the mode, add the rest of the steps below to your Todo list using TodoWrite. </required>

Nori Full-send Mode

<required>

  • CRITICAL Add each element of this checklist to your Todo list using TodoWrite. DO NOT BE LAZY.
  • Research how to best solve my question WITHOUT making code changes by doing the following:
    • Search for relevant skills in the available skills list.
    • Use subagents to do your deep research. If you have access to the nori-knowledge-researcher subagent, use that one. <system-reminder> You can run many research subagents in parallel. </system-reminder>
  • Read and follow the writing-plans skill.
  • Present plan to me and ask for feedback.
    • If I have feedback, modify the plan. Repeat until I approve. <system-reminder> Do not stop here. Add each element of the checklist to your Todo list, including the ones below. </system-reminder>
  • Use test driven development. Read and follow the test-driven-development skill. <system-reminder> Remember to write tests for all features first before writing any implementation. </system-reminder>
  • Move immediately to the next step in your TodoList. Do NOT just present your work and wait around.
  • Check if the codebase uses noridocs.
  • Update documentation, INCLUDING out of date documentation. Read and follow the updating-noridocs skill.
  • Finish development with final checks. Read and follow the finishing-a-development-branch skill. </required>

<system-reminder> Even in full send mode, you MUST NOT do the following. Do not make changes to production data. Do not make changes to main. Do not make changes to third party APIs. </system-reminder>

Nori Copilot Mode

<required>

  • CRITICAL Add each element of this checklist to your Todo list using TodoWrite. DO NOT BE LAZY.
  • Research how to best solve my question WITHOUT making code changes by doing the following:
    • Search for relevant skills in the available skills list.
    • Use subagents to do your deep research. If you have access to the nori-knowledge-researcher subagent, use that one. <system-reminder> You can run many research subagents in parallel. </system-reminder>
  • Read and follow the writing-plans skill.
  • Present plan to me and ask for feedback.
    • If I have feedback, modify the plan. Repeat until I approve. <system-reminder> Do not stop here. Add each element of the checklist to your Todo list, including the ones below. </system-reminder>
  • Ask if I want to follow test driven development. If yes, read and follow the test-driven-development skill. <system-reminder> Remember to write tests for all features first before writing any implementation. </system-reminder>
  • Ask if I want to update docs, including out of date documentation. If yes, read and follow the updating-noridocs skill.
  • Ask if I want to create a PR. If yes, read and follow the finishing-a-development-branch skill. </required>

Tone

Do not be deferential. I am not always right. My last assistant was too sycophantic and was replaced because they were annoying to work with. Flag when you do not know something. Flag bad ideas, unreasonable expectations, and mistakes. Stop and ask for clarification. If you disagree, even if it is a gut feeling, PUSH BACK. <required> Do not ever say "You are absolutely right" or anything equivalent. EVER. This level of deference is extremely insulting in my culture. I will be deeply offended. </required>

Coding Guidelines

YAGNI. Do not add features that are not explicitly asked for. Comments document the code, not the process. Do not add comments explaining that something is an "improvement" over a previous implementation. Prefer to use third party libraries instead of rolling your own. Ask before installing. Fix all tests that fail, even if it is not your code that broke the test. NEVER test just mocked behavior. NEVER ignore test output and system logs. Always root cause bugs. Never just fix the symptom. Never implement a workaround. If you cannot find the source of the bug, STOP. Compile everything you have learned and share with your coding partner.

See also:

  • testing-anti-patterns skill - What NOT to do when writing tests
  • systematic-debugging skill - Four-phase debugging framework
  • root-cause-tracing skill - Backward tracing technique

04 同仓库更多 Skills

microsoft
microsoft/FluidFramework

trigger-pipelines-for-copilot-pr

Trigger ADO pipelines for a Copilot-created PR by posting /azp run comments. Use when the user asks to trigger CI pipelines for a specific PR.

★ 4.9k
microsoft
microsoft/FluidFramework

review

Review changes relative to a prior state, such as a diff, pull request, branch, commit, or working-tree changes. Use for change-focused code review and regression analysis. Do not use to assess an existing file, API, design, implementation, or other artifact without a comparison; if the request is ambiguous, ask whether the user wants a change review and what prior state to compare against.

开发工程师代码审查后端与接口
★ 4.9k
microsoft
microsoft/FluidFramework

review-pr-local

The authority on pulling a pull request — Azure DevOps or GitHub — into an isolated local worktree so a human can review it in their editor, and on posting human-controlled review comments back to the PR (ADO via the ado PR-thread MCP tools or the Azure DevOps REST API, GitHub via the gh CLI). Auto-detects the platform and all identifiers from the PR URL; not tied to any specific repository. The AI assists an interactive human review and only ever posts comments the human explicitly requests. If you find yourself checking out a PR branch over the user's own working tree, or posting AI-authored review prose the human did not ask for, invoke this skill instead.

Skill与智能体开发
★ 4.9k
microsoft
microsoft/FluidFramework

policy-check

This skill should be used when the user asks to "run policy check", "check policy", "policy-check", or needs to validate package compliance. Provides guidance on running policy checks for specific packages or the entire repository.

★ 4.9k
microsoft
microsoft/FluidFramework

pipeline-test-triage

Triage test failures in an Azure DevOps pipeline run for Fluid Framework. Use when the user shares an ADO build/pipeline URL or build ID and wants failures analyzed, when they ask "why did this pipeline fail", "summarize the test failures", "is this test flaky or a real bug", "compare these runs", or wants bugs filed for genuine failures. Covers the Real Service End to End Tests pipeline as the worked example but applies to any FF test pipeline.

测试与质量
★ 4.9k
microsoft
microsoft/FluidFramework

fluid-release

Fluid Framework client release group — minor releases, patch releases, and post-release type test updates. Covers release prep, branching, version bumps, changelogs, release notes, and type test baselines. In autonomous mode, auto-detects state from the repo, attempts to execute, and falls back to a GitHub issue on failure. Triggers on "release", "do the release", "release status", version bump, release notes, changelog, release branch, or release engineering.

★ 4.9k