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
- 所属仓库
- emo-eth/skills
- 最近更新
- 2026-11-13
- 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 5
- 许可证LICENSE1.0 KB
- 参考文档references/divergent-reviewer.md3.8 KB
- 参考文档references/judgment-reviewer.md3.2 KB
- 参考文档references/synthesizer.md4.1 KB
- 参考文档references/tooling-reviewer.md4.6 KB
03 SKILL.md 正文
Reflect
Read pstack-runtime before the first host-specific operation.
Mine the current conversation for durable learnings, then route them into skill edits.
When to invoke
- The user said "reflect" or "/reflect".
- A complex task (5+ tool calls) just landed cleanly and the recipe is worth keeping.
- The agent hit dead ends, found the working path, and the path generalizes.
- The user corrected the agent's approach mid-task.
- A non-trivial workflow emerged that isn't captured anywhere.
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
Use agent-conv thread <current-workspace> --source omp when available. The exact current directory is the workspace query. Otherwise inspect only the matching workspace under ~/.omp/agent/sessions/, order candidates by modification time, and match the opening user prompt. If no path resolves, write a tight session digest.
2. Spawn three reviewers in parallel
Launch three read-only reviewers in one OMP task batch. Read reflect-judgment and reflect-tooling from ~/.config/pstack/omp-agents.json when present. Otherwise use reviewer for all three. The prompts forbid file writes; the parent applies approved edits.
| Lens | OMP agent 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 the task result.
3. Synthesize
Spawn one read-only synthesis task. Use reflect-judgment from the OMP config, or reviewer by default. Use references/synthesizer.md verbatim and pass large reviewer outputs through local:// artifacts. The synthesizer returns Accepted / Rejected / Backlog.
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. The synthesizer already applies this criterion; this is a final pass before edits land. See the encode-lessons-in-structure principle 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 in the org; do not auto-apply.
Report Backlog items. File one only when the user asked for tracker updates and the tracker tool is available.
For each approved Accepted item, follow the Routing field exactly:
- Trivial existing-skill edit: apply it directly in
~/dev/skills/skills/<name>/. - Substantive edit, trigger tuning, or new skill: follow
skill://poteto-mode/playbooks/authoring-a-skill.mdin~/dev/skills. - After any approved skill edit, follow
~/dev/skills/AGENTS.md: update project state when needed, commit, pushmain, and refresh the installed skill withnpx skills.
Validate every changed skill by checking its frontmatter, folder/name match, relative links, and OMP-only tool vocabulary.
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 reported or filed:
<issue title>(<tags>). One line each. - Dropped: one line per rejected finding + reason from the synthesizer.
04 同仓库更多 Skills
system-vibe
Personal operating-system contract for the home fleet, sync, SSH, skills, Herdr, and local inference. Use when diagnosing or changing machines, chezmoi, SSH, skill distribution, Windows sleep/wake, or the local model router. This is the qualitative list of what must stay true; agents invent enforcement from the clauses, not from tribal setup lore.
rubber-stamp-travel-field-note
Use for rubber-stamp travel field-note photo posters.
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. Must always apply.
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.