菜单
Use when preparing notes for a 1:1 with a report, drafting written feedback, or summarizing a person's recent work/contributions ahead of a conversation or review cycle.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when asked to generate or update architecture diagrams or ER (entity-relationship) diagrams from the codebase — the code is the source of truth, not existing docs.
Use when the user asks how to reduce Claude Code costs, manage token usage, choose the right model for a task, or wants a review of whether a session/workflow is running efficiently.
Use when the user asks for a standup update, daily summary, "what did I work on yesterday", or wants to recap recent commits before a team sync.
Use when investigating a bug, test failure, or unexpected behavior and proposing a fix — always runs investigation in Plan Mode so no code changes happen until the root cause and approach are confirmed.
Use when writing or reviewing code that crosses a trust boundary (user input, external APIs, file/network/DB I/O) — guards against invalid input, injection, unsafe failure modes, and missing validation at the edges without over-defending internal code.
Use when reviewing local code changes for correctness and assessing the blast radius / impact of a change — what else could break, who/what depends on the changed code.
| name | 1on1-prep |
| description | Use when preparing notes for a 1:1 with a report, drafting written feedback, or summarizing a person's recent work/contributions ahead of a conversation or review cycle. |
Helps an EM walk into a 1:1 or write feedback prepared — grounded in actual recent work, not vague impressions. This skill drafts prompts for your own thinking and conversation, not verdicts about a person.
Pull recent, concrete signal about the person's work (with their knowledge — this is prep for a conversation with them, not surveillance):
git log --author=...), issues they closed/are
working on (via the connected Jira/Linear MCP), code review comments
they gave or receivedjira-progress-pulse history if available)Turn that into talking points, not a performance review:
## 1:1 prep — <name> — <date>
**Recent work worth acknowledging**
- <specific, concrete — "shipped X" not "has been productive">
**Things to ask about**
- <something that looked stuck, ambiguous, or like it might need support —
framed as a question, not a conclusion>
**Career/growth thread to follow up on**
- <whatever's open from the last conversation — don't let this drop>
Ask what the feedback is for (review cycle? in-the-moment course correction? recognition?) — the right tone and specificity depend entirely on this, and getting it wrong is worse than not drafting at all.
Ground every point in a specific, verifiable instance — "in the auth refactor PR, you caught an edge case the rest of us missed" beats "good attention to detail." Vague feedback is unfalsifiable and unhelpful.
Draft, then explicitly hand it back for the EM to apply their own judgment about tone, relationship history, and timing — things no amount of context-gathering substitutes for. This is the clearest case in this whole toolkit where your domain knowledge — about this specific person, this relationship, this moment — has to be the final filter.