一键导入
maestronote
Capture decisions, constraints, and context to persistent notepad. Priority notes are injected into every session and implementation run.
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。
菜单
Capture decisions, constraints, and context to persistent notepad. Priority notes are injected into every session and implementation run.
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 bootstrapping, updating, or reviewing AGENTS.md — teaches what makes effective agent memory, how to structure sections, signal vs noise filtering, and when to prune stale entries
Use before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes
Deep discovery and specification for ambitious features. Full BMAD-inspired interview with classification, vision, journeys, domain analysis, and FR synthesis. Same output contract (spec.md + plan.md) as a standard feature but far richer. Use for multi-component systems, regulated domains, or unclear requirements.
Use when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies
Use when working with Docker containers — debugging container failures, writing Dockerfiles, docker-compose for integration tests, image optimization, or deploying containerized applications
| name | maestro:note |
| description | Capture decisions, constraints, and context to persistent notepad. Priority notes are injected into every session and implementation run. |
| argument-hint | <content> [--priority|--manual|--show|--prune|--clear <section>] |
Notes are the cross-cutting memory layer. They persist across sessions, survive context resets, and -- for Priority Context -- are automatically injected into every worker prompt.
Core principle: If a future session needs this to avoid a mistake, capture it. If not, skip it.
Two-tier system:
Always:
Never:
task-updatecontext-write for codebase knowledgeDecision heuristic: "Will a future session need this to avoid a mistake?" Yes = capture. No = skip.
| Attribute | Priority Context | Working Memory |
|---|---|---|
| Injection | Every session + every worker prompt | On-demand (show/prune) |
| Audience | All agents, including workers who cannot ask | Orchestrator only |
| Content | Hard constraints, blocking decisions | Insights, learnings, soft preferences |
| Lifespan | Until feature ships or constraint lifts | Until pruned |
| Volume | 3-7 bullets max (more = noise) | Unlimited (prune periodically) |
| Format | Imperative constraint | Dated insight with track context |
Promoting to Priority: If a Working Memory note keeps causing mistakes across sessions, promote it to Priority Context. If a Priority note no longer applies, demote or delete it.
See reference/note-patterns.md for more examples across decisions, constraints, discoveries, and dependencies.
$ARGUMENTS
| Input | Target Section |
|---|---|
<content> (no flag) | Working Memory |
--priority <content> | Priority Context |
--manual <content> | Manual |
--show | Display full notepad (read-only) |
--prune | Remove stale Working Memory entries |
--clear <section> | Clear a section (priority, working, all) |
If .maestro/notepad.md does not exist, create it with section headers:
# Notepad
## Priority Context
## Working Memory
## Manual
Add (default, --priority, --manual):
.maestro/notepad.md- <content> after the header, before the next ##Show (--show):
.maestro/notepad.md/maestro:note <content> to start."Prune (--prune):
## Working Memory## Priority Context and ## Manual intactClear (--clear <section>):
priority, working, or allCreate --> Review (session start) --> Update (constraint changes) --> Archive/Delete (resolved)
| Trigger | Action |
|---|---|
| New decision or constraint discovered | Create note (choose tier) |
| Session start | Review Priority Context -- still accurate? |
| Constraint changed | Update the note in place (don't append a duplicate) |
| Feature shipped | Prune Working Memory, demote/delete Priority notes |
| Constraint lifted | Delete the Priority note |
| Reasoning might matter later | Archive to context-write before deleting |
| Anti-Pattern | Why It Hurts | Fix |
|---|---|---|
| Notes too vague ("the API thing") | Future sessions cannot act on it | Include specifics: which API, what behavior, what constraint |
| Notes too long (full paragraphs) | Agents skim or skip long bullets | One line per note. Details go in context-write |
| Duplicating spec/plan content | Two sources of truth that drift apart | Reference the spec; don't copy it |
| Priority Context bloat (>7 items) | Agents stop reading; noise drowns signal | Prune aggressively. Only hard constraints belong here |
| Never pruning Working Memory | Graveyard of stale insights misleads agents | Prune after each feature completion |
| Using notes for codebase knowledge | Notes are ephemeral memory, not documentation | Use context-write for durable codebase knowledge |
| Status updates as notes | Notes are for decisions and constraints, not progress | Use task-update for status |
| Section | Written By | Read By | Persistence |
|---|---|---|---|
| Priority Context | User via --priority | maestro:implement Step 3.8, team-mode worker prompts | Until manually cleared |
| Working Memory | Default /maestro:note, auto-capture during implementation | Sessions, prune | Pruned periodically |
| Manual | User via --manual | Sessions | Until manually cleared |
/maestro:setup -- Scaffold project context (run first)/maestro:new-track -- Create a feature/bug track with spec and plan/maestro:implement -- Execute implementation (reads Priority Context at Step 3.8)/maestro:review -- Verify implementation correctness/maestro:status -- Check progress across all tracks/maestro:note -- You are here. Capture decisions and context to persistent notepadPriority Context is automatically loaded by /maestro:implement at execution start and injected into worker prompts in team mode. Working Memory accumulates insights from both manual notes and auto-capture during implementation. Use --prune after each feature completion to keep Working Memory relevant.