Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/hack-dance/hack/control-plane-workernpx skills add hack-dance/hack --skill control-plane-workergit clone --depth 1 https://github.com/hack-dance/hackWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00026 | $0.01310 |
| Opus 5 | $0.00013 | $0.00655 |
| Sonnet 5 | $0.00005 | $0.00262 |
| Haiku 4.5 | $0.00003 | $0.00131 |
Grade A, and why
control-plane-worker scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 93 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Control Plane Worker
NOTE: Startup and cleanup are handled by worker-base. This skill defines the WORK PROCEDURE.
When to Use This Skill
Use this skill for features that primarily touch:
src/**CLI and control-plane code.hack/docker-compose.yml,.hack/hack.config.json, or other source-of-truth Hack runtime files- local runtime orchestration, env/runtime hardening, lifecycle processes, tickets, sessions, MCP/agent setup, docs, or the slim macOS companion
Do not use this skill for retired v3 surfaces:
- hosted auth/account/org/team management
- web dashboard work
- built-in GitHub workflows
- built-in Linear sync/project-artifact flows
Required Skills
hack-cli— invoke when the feature touches.hack/**, runtime orchestration, lifecycle/session flows, tickets, env, or anyhack up/ps/open/downverification.
Work Procedure
- Read the assigned feature,
mission.md, missionAGENTS.md,.factory/services.yaml, and relevant.factory/library/*.mdfiles. Restate the exact assertions or outcomes the feature must complete. - Investigate existing code paths and add the failing test or regression harness first. Prefer the narrowest relevant suites under
tests/*.test.ts. If the feature has nofulfillsclaims, still add characterization or regression coverage for the changed behavior. - Implement the smallest coherent change set in CLI, runtime config, tickets, env, lifecycle, macOS, or agent setup. Never hand-edit
.hack/.internal/**or.hack/.branch/**; only change source-of-truth files. - Run focused validators first, then the smallest meaningful
typecheck/checkcommands for the touched surfaces. For repo-bound CLI behavior, build and validate with./dist/hackor repo-local Bun entrypoints. When invokingbun testfrom the repo root against files outside./tests, use absolute paths or explicit./-prefixed paths that Bun actually honors in this repo so targeted commands do not silently skip files.- If the assigned feature is explicitly about fixing a known red baseline, capture the failing baseline evidence once, then continue the repair work and rerun the gate before handoff.
- If repo-bound GitHub CLI routes cannot reach the changed auth code because
dance.hack.githubis not enabled in project config yet, use a direct resolver or similarly narrow deterministic smoke and record why the repo-bound path was unavailable. - If no safe repo-bound hook exists to force a failure mode (for example local-sync failure injection), deterministic regression tests are acceptable proof as long as you explain why a live manual repro would mutate real project state.
- For daemon/gateway request-target hardening, raw-socket regression coverage against the proxy transport is preferred. If you also need live proof without mutating shared user daemon state, an isolated temp-HOME
bun index.ts daemon start --foregroundsmoke is an acceptable validation pattern; record the isolation setup in the handoff. - For lifecycle changes, verify shell semantics, process-group cleanup, stale pane/process metadata reconciliation, singleton listener behavior, and doctor recovery guidance.
- For env changes, verify overlay order, worktree-local override behavior, linked-worktree secret-key lookup, host-vs-compose target behavior, and materialization drift detection.
- Capture any blockers, discovered issues, or scope mismatches immediately. If a feature needs credentials, unavailable infrastructure, or a change that would reintroduce hosted/web/integration dependencies, return to the orchestrator instead of guessing.
- Stop any processes you started and produce a detailed handoff with exact commands, observations, tests added, and remaining issues.
- If
bun run checksucceeds and only re-surfaces the known warning-only complexity diagnostics already documented in missionAGENTS.md, do not return to the orchestrator for that reason and do not record them as new discovered issues unless your feature directly worsened the warned files.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 93 lines · 26 tokens per session scan A c0b16baac557
control-plane-worker is a skill published in the GitHub repository hack-dance/hack (8 stars, last pushed 20d ago), licensed MIT. It adds 26 tokens to every session and 1,310 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…