AGENT SKILL

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
2026-11-14
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.

02 资源文件 4

03 SKILL.md 正文

Reflect

Mine the current conversation for durable learnings, then route them into skill edits.

When to invoke

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.

Process

1. Locate the active transcript

The parent finds its own transcript file before fanning out. Claude Code keeps each session at ${CLAUDE_CONFIG_DIR:-~/.claude}/projects/<slug>/<session-id>.jsonl, where <slug> is the workspace path with every / turned into - (so /Users/you/proj becomes -Users-you-proj), and its subagents under <session-id>/subagents/. Codex keeps sessions at ~/.codex/sessions/<yyyy>/<mm>/<dd>/rollout-*.jsonl. Every line is one event. Look only in the current workspace's directory. Globbing across projects reads private chats from unrelated projects.

ls -t "${CLAUDE_CONFIG_DIR:-$HOME/.claude}/projects/<slug>"/*.jsonl | head -5

The newest file is usually the active session. Confirm by grepping it for the conversation's opening user prompt. If no path resolves, write a tight digest of the session and pass that instead.

2. Spawn three reviewers in parallel

One message, three subagents. Reviewers need MCP access for context lookups (tickets and chat threads referenced in the transcript), so don't use a read-only agent type that strips it. If another model family is available, give it the tooling lens.

Lens Prompt template
Judgment references/judgment-reviewer.md
Tooling references/tooling-reviewer.md
Divergent 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 subagent with MCP access, since its quality check spot-verifies citations. Use references/synthesizer.md verbatim, with each reviewer's full output inlined where marked. The synthesizer returns a structured Accepted / Rejected / Backlog list.

4. Structural enforcement check

Sanity-check the synthesizer's Accepted list. For any item that would be enforced more reliably by a lint rule, script, metadata flag, or runtime check, move it from Accepted to Backlog. See the principle-encode-lessons-in-structure skill.

5. Apply

Before applying any Accepted edit, present the synthesizer's full Accepted/Rejected/Backlog output to the user and wait for explicit approval. The user picks which subset to apply and may redirect routings. Skill changes affect every future agent. Do not auto-apply.

List Backlog items in the reply. File them to a tracker only if the user asks.

For each approved Accepted item, follow the Routing field exactly:

  • Trivial existing-skill edit (a one-line bullet, a tightened sentence, a stale fact corrected): parent does directly.
  • Substantive existing-skill edit (a new section, a new pattern table, more than ~10 lines): hand to the skill-creator skill and run its draft / test / iterate loop.
  • tune description: <skill path> (the skill exists but didn't trigger when it should have): hand to skill-creator and run its description-optimization loop.
  • new skill: <kebab-name>: hand creation to skill-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. Skip this step if it doesn't.

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: one line each.
  • Dropped: one line per rejected finding + reason from the synthesizer.

04 同仓库更多 Skills

stepankuzmin
stepankuzmin/svenska-app

code-review

暂无简介

★ 0
stepankuzmin
stepankuzmin/svenska-app

wrangler

Run or troubleshoot Wrangler CLI commands in projects that have a wrangler.jsonc, wrangler.json, or wrangler.toml file, including local development, Previews, deployment, and resource management. For new projects and projects without a Wrangler configuration file, use the cf CLI instead.

★ 0
stepankuzmin
stepankuzmin/svenska-app

workers-best-practices

Cloudflare Workers best practices for production applications. Use when writing, reviewing, or configuring Workers.

★ 0
stepankuzmin
stepankuzmin/svenska-app

why

Use when the user asks to investigate why code is built the way it is: 'why does X work this way', 'why we picked Y', design rationale, regressions, or postmortems. Not for explaining a single error or a quick question. Queries each available evidence source (git and PRs, issues, docs, team chat) in parallel, then returns a cited read on decisions and tradeoffs. Use how for runtime behavior.

★ 0
stepankuzmin
stepankuzmin/svenska-app

vercel-react-best-practices

React and Next.js performance optimization guidelines from Vercel Engineering. This skill should be used when writing, reviewing, or refactoring React/Next.js code to ensure optimal performance patterns. Triggers on tasks involving React components, Next.js pages, data fetching, bundle optimization, or performance improvements.

★ 0
stepankuzmin
stepankuzmin/svenska-app

vendor-skill

Adapt an external skill into this marketplace, or review upstream changes to one already adapted. Use when asked to vendor, import, adapt, mirror, or update a skill from skills.sh, `npx skills add`, or any external GitHub skills repo, especially to work around Claude Code Web only allowing one marketplace per GitHub repo name.

★ 0