Architecture and protocols for a Level 7.5 project manager — a Python script that coordinates multiple Claude Code worker sessions. Use when designing, building, or operating a PM that sits between a human director and a swarm of CC workers. Covers escalation engine, worker lifecycle, feedback translation, coherence monitoring, and demo generation.
Coordinates multiple Claude Code workers with deterministic orchestration and AI-driven escalation for complex projects.
npx claudepluginhub rbergman/dark-matter-marketplaceThis skill inherits all available tools. When active, it can use any tool Claude has access to.
A Python script (not a CC session) that sits between a human director and a swarm of Claude Code worker sessions. Uses the Claude Agent SDK for targeted LLM calls at decision points while handling bookkeeping deterministically.
Human Director → NL Frontend → PM Script → Worker Sessions (in worktrees)
The PM uses beads for work tracking and git worktrees for worker isolation.
| Subsystem | Purpose | Reference section |
|---|---|---|
| Escalation engine | 3-tier (Log/Notify/Block) decisions across 12 domains | references/pm-script-architecture.md §2 |
| Worker lifecycle | Spawn → claim → checkpoint → rotate → complete | §3 |
| Feedback translation | Human intent → worker-scoped instructions | §4 |
| Demo/coherence | Progress summaries, architectural drift detection | §5, §6 |
| Spinning detection | Identify stuck workers (3+ attempts, no progress) | §3 |
Before implementing any PM subsystem, read the relevant reference:
references/pm-script-architecture.md — Full architecture: escalation taxonomy, worker state machines, feedback protocols, coherence monitoring, demo generation. (~800 lines)references/piloting-guide.md — Phased adoption: single worker MVP → parallel workers → NL frontend → coherence monitoring. Start here for first-time setup.The architecture doc is large. Use targeted reads:
grep -n "^## " references/pm-script-architecture.md # Section index
grep -n "<topic>" references/pm-script-architecture.md # Find specific content
| Need | Tier | Tool |
|---|---|---|
| Focused single-task delegation | L5 | dm-work:orchestrator with Task() subagents |
| Multi-agent discussion/debate | L7 | dm-team:lead with Agent Teams |
| Multi-session worker coordination | L7.5 | PM Script (this skill) |
| Full platform orchestration | L10 | Gastown |
Choose L7.5 when you need workers with independent context windows, session rotation, and deterministic coordination — but don't need a full platform.
Expert guidance for Next.js Cache Components and Partial Prerendering (PPR). **PROACTIVE ACTIVATION**: Use this skill automatically when working in Next.js projects that have `cacheComponents: true` in their next.config.ts/next.config.js. When this config is detected, proactively apply Cache Components patterns and best practices to all React Server Component implementations. **DETECTION**: At the start of a session in a Next.js project, check for `cacheComponents: true` in next.config. If enabled, this skill's patterns should guide all component authoring, data fetching, and caching decisions. **USE CASES**: Implementing 'use cache' directive, configuring cache lifetimes with cacheLife(), tagging cached data with cacheTag(), invalidating caches with updateTag()/revalidateTag(), optimizing static vs dynamic content boundaries, debugging cache issues, and reviewing Cache Component implementations.
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.