maestro-companion
Quick execution for small tasks — minimal run lifecycle (start + done) with evidence recording. Full LLM capability, scoped to mechanically clear tasks.
Intent-driven knowledge-store and Run knowledge lifecycle management — read-only audit/prune reports, stage candidates (with signal recording), review/resolve/promote candidates, harvest artifacts, or manage wiki/domain knowledge.
$ npx -y skills add catlog22/maestro-flow --agent claude-codeHow it fires
How this command gets triggered: by you, by Claude, or both.
/maestro-knowledgeContext preview
What this command does when you run it.
Intent-driven knowledge-store and Run knowledge lifecycle management — read-only audit/prune reports, stage candidates (with signal recording), review/resolve/promote candidates, harvest artifacts, or manage wiki/domain knowledge.
name: maestro-knowledge disable-model-invocation: true description: Intent-driven knowledge-store and Run knowledge lifecycle management — read-only audit/prune reports, stage candidates (with signal recording), review/resolve/promote candidates, harvest artifacts, or manage wiki/domain knowledge. argument-hint: "[intent — e.g. '审计知识库' | 'harvest 这个 session' | 'wiki health' | '注册术语 MVP' | 'extractors']" allowed-tools: - Read - Write - Edit - Bash - Glob - Grep - Agent - WebFetch - AskUserQuestion session-mode: none
<purpose> Intent-driven knowledge-store management. No fixed grammar — state your intent; the command classifies it and runs the matching workflow or direct lifecycle command. Explicit keywords still work as deterministic shortcuts.
| Operation | Keywords | Execution document or CLI | |-----------|----------|--------------------------| | audit | `audit` / 审计 / 检查知识库 | `maestro knowledge audit --scope all`(只读;细节见 `~/.maestro/workflows/knowledge-audit.md`) | | normalize | `normalize` / `migrate` / 规范化 / 迁移 | `maestro knowledge normalize --report <path>`,审阅后才可原样追加 `--apply` | | review | `review` / 审查 / 证据 / 下一步 / 匹配 / 去重 / 冲突检测 / 裁决 / 候选 / backlog | `maestro knowledge review <session-id> [--refresh] [--resolve <id> --as <choice> --reason "..."]` | | stage | `stage` / 暂存 / candidate / 沉淀候选 / cited / validated / contradicted / 记录命中关系 | `maestro knowledge stage ... [--signal <signal> --signal-ids <ids>]` | | promote | `promote` / 晋升 / 发布候选 | `maestro knowledge promote ... [--all]` | | harvest | `harvest` / 提取 / 收割 / 从工件 | `~/.maestro/workflows/harvest.md` | | wiki | `wiki` / 知识图谱 / 连接 / 摘要 / 健康 | `~/.maestro/workflows/wiki-manage.md` / `~/.maestro/workflows/wiki-connect.md` / `~/.maestro/workflows/wiki-digest.md` | | extractors | `extractors` / 抽取器 / 生成抽取规则 | `~/.maestro/workflows/extractors.md` | | domain | `domain` / 领域术语 / 注册术语 / term | `~/.maestro/workflows/domain-add.md` | </purpose>
<dispatch> Classify the intent in `$ARGUMENTS` into one operation. For an operation mapped to an execution document, read the path shown in the table directly and follow it; do not create a Session or Run merely to load instructions. For direct lifecycle operations, invoke the listed `maestro knowledge` CLI command.
1. Explicit keyword present → use its execution document or direct CLI lifecycle command (deterministic shortcut). 2. Otherwise infer from the intent (see the table above), e.g. "审计/检查知识库" → audit, "从工件/session 提取" → harvest, "知识图谱/wiki 健康" → wiki, "注册术语 X" → domain. 3. `review` / `stage` / `promote` map directly to the corresponding `maestro knowledge` CLI. `review --refresh` includes reconciliation; `review --resolve` includes disposition resolution; `stage --signal --signal-ids` includes signal recording. Preserve stable knowledge IDs, graph aliases, Run ID, Session ID, signal, candidate ID, disposition, target, and reason exactly; do not translate these operations into direct spec/knowhow writes. 4. For wiki, classify the sub-action: `connect`/连接 → `~/.maestro/workflows/wiki-connect.md`; `digest`/摘要 → `~/.maestro/workflows/wiki-digest.md`; `health`/`search`/`cleanup`/`stats`/健康/检查/_(none)_ → `~/.maestro/workflows/wiki-manage.md`. 5. Ambiguous → display the operation table and ask the user to pick.
Intent-driven workflow orchestration for multi-agent AI development — adaptive lifecycle engine, self-reinforcing knowledge graph, and visual dashboard for Claude Code, Gemini, Codex & more
Repo: catlog22/maestro-flow
Quick execution for small tasks — minimal run lifecycle (start + done) with evidence recording. Full LLM capability, scoped to mechanically clear tasks.
Use when designing, reviewing, refining, fixing, or codifying frontend UI with Maestro's self-contained Impeccable core
Intent-driven issue lifecycle management — describe what you want in natural language (报告一个 bug / 列出开放 issue / 关掉 ISS-xxx / 关联到 task / 扫描发现问题) and the workflow…