reflect
Spawn three parallel review subagents over the active transcript, surface learnings, and route each to a concrete edit on an existing skill. Use when the user says reflect.
- License 许可
- MIT
- 最近更新
- 2026-10-30
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- reflect
- description
- Spawn three parallel review subagents over the active transcript, surface learnings, and route each to a concrete edit on an existing skill. Use when the user says reflect.
- disable-model-invocation
- true
02 资源文件 4
- 参考文档references/divergent-reviewer.md3.6 KB
- 参考文档references/judgment-reviewer.md3.1 KB
- 参考文档references/synthesizer.md4.0 KB
- 参考文档references/tooling-reviewer.md4.4 KB
03 SKILL.md 正文
Reflect
Mine the current conversation for durable learnings, then route them into skill edits.
Invoke when the user says "reflect" or "/reflect". Skip when the conversation is trivial, off-topic, or already covered by an existing skill the parent followed correctly. One-offs are not learnings.
1. Locate the active transcript
Before fanning out, find this conversation's transcript. In Claude Code it lives in ~/.claude/projects/<slug>/, where <slug> is the working directory with every / replaced by -. Do not glob across ~/.claude/projects/*/. That crosses workspace boundaries and reads private chats from unrelated projects. Codex keeps every project's sessions under ~/.codex/sessions/YYYY/MM/DD/, so check that the first line's payload.cwd is the working directory before reading further. Grok keeps ~/.grok/sessions/<encoded-cwd>/<session-id>/chat_history.jsonl, one directory per workspace, and grok sessions list names this workspace's sessions.
ls -t ~/.claude/projects/<slug>/*.jsonl ~/.claude/projects/<slug>/*/subagents/*.jsonl 2>/dev/null | head -10 # Claude Code
ls -t ~/.codex/sessions/*/*/*/*.jsonl 2>/dev/null | head -10 # Codex
grok sessions list # Grok
That covers the Claude Code session (<id>.jsonl) and subagent (<id>/subagents/<child>.jsonl) layouts, and Codex's dated rollout files. For each candidate, grep -l a distinctive phrase from the conversation's opening user prompt. Take the file that contains it. If none matches, pass a tight digest of the session instead.
2. Spawn three reviewers in parallel
Spawn three general-purpose subagents at once, not read-only. Reviewers need MCP access to look up the tickets, chat threads, and observability traces the transcript references, and read-only modes can block MCP tools.
Each reviewer and the synthesizer name a role in ~/.config/agent-workflows/models.toml. Resolve its model with aw-run --resolve <role> and spawn per the Subagents section of the workflow skill. A role of inherit runs on the parent model. If the spawn or aw-run rejects the model, use the model's fallback and say so. Pick each subagent's reasoning level and fast or standard speed for the job, per the Subagents section of the workflow skill.
| Lens | Role | Prompt template |
|---|---|---|
| Judgment | reflect_judgment |
references/judgment-reviewer.md |
| Tooling | reflect_tooling |
references/tooling-reviewer.md |
| Divergent | reflect_judgment |
references/divergent-reviewer.md |
Pass each template verbatim, substituting the transcript path or digest where marked. Reviewers return findings in their final message.
3. Synthesize
One general-purpose subagent on the reflect_judgment role, not read-only. It spot-verifies citations through MCP, and read-only modes can block MCP tools. Pass references/synthesizer.md verbatim, with each reviewer's full output inlined where marked. It returns an Accepted / Rejected / Backlog list.
4. Structural enforcement check
Move any Accepted item that a lint rule, script, metadata flag, or runtime check would enforce more reliably to Backlog. See the encode-lessons-in-structure principle skill.
5. Apply
Present the synthesizer's full Accepted / Rejected / Backlog output and wait for explicit approval before applying any Accepted edit. The user picks the subset and may redirect routings. Skill changes affect every future agent in the org. Do not auto-apply.
File each Backlog item to your team's devex or backlog tracker without waiting. Only the Accepted list waits for approval.
Follow each approved row's Routing exactly:
- Trivial existing-skill edit (a one-line bullet, a tightened sentence, a stale fact corrected): the parent does it directly.
- Substantive existing-skill edit (a new section, a new pattern table, more than ~10 lines): hand to the CLI's
skill-creatorskill and run its draft / test / iterate loop. tune description: <skill path>(the skill exists but didn't trigger when it should have): hand toskill-creatorand run its description-optimization loop.new skill via skill-creator: <kebab-name>: hand creation toskill-creator. Do not invent the shape ad hoc.
If your environment ships a SKILL.md validator, run it on every touched skill before declaring done.
6. Summarize for the user
Short list, no preamble:
- Edits applied:
<skill path>. What changed, one line each. - New skills created:
<skill path>. One line each (rare). - Backlog filed to the devex tracker:
<issue title>(<tags>). One line each. - Dropped: one line per rejected finding + reason from the synthesizer.
04 同仓库更多 Skills
workflow
Concise, detailed responses, deliberate subagents, unslopped prose, simple code, and verified work, with each task routed to a playbook. Use for /workflow, or for requests to work in this style. `aw skills always workflow` loads it into every session.
why
Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed thresholds. Discovers available MCPs and queries each evidence category (source control, issue tracker, long-form docs, real-time chat, infrastructure observability, error tracking, product analytics warehouse) in parallel, then returns a cited read on decisions and tradeoffs. Use how for runtime behavior.
unslop
Cut AI tells from any writing.
typescript-best-practices
TypeScript best practices. Use when reading or editing any .ts or .tsx file.
technical-writing
Layered technical-writing standard: Diátaxis structure, Google developer style sentences, STE instruction rules, Global English syntax. Use for /technical-writing or when writing or reviewing docs, RFCs, readmes, PR descriptions, or commit messages.
teach
Explain a body of work plainly so a person actually understands it. Runs the `how` and `why` skills and weaves what they find into one clear explanation. Use for 'teach me this', 'help me really understand X', 'explain this change or subsystem to me'.