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/rishi4792/compass/startnpx skills add Rishi4792/compass --skill startgit clone --depth 1 https://github.com/Rishi4792/compassWhat 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 | $0.00093 | $0.03831 |
| Opus 5 | $0.00046 | $0.01916 |
| Sonnet 5 | $0.00019 | $0.00766 |
| Haiku 4.5 | $0.00009 | $0.00383 |
Grade A, and why
start 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 2d ago.
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 — 114 lines — stays where its author put it; the contents beside it link to each section on GitHub.
compass:start — the lifecycle orchestrator
Compass builds software true to spec, with zero drift. The contract is the invariant; every stage is checked against it. This skill sequences the stage skills, but never auto-advances in gated mode — each transition is a user-driven gate, AND each downstream stage runs a real script gate that blocks on missing proof.
The whole lifecycle is driven through the single front door
/compass:go, which reads state and routes into each stage skill (compass:contract,compass:review-contract,compass:plan,compass:review-plan,compass:build,compass:review-build,compass:ship). Resume from a fresh terminal with/compass:resume. The per-stage skills are hidden from the/menu (user-invocable: false) but Compass invokes them for you — you never have to remember which one.
Read these before you start. They are the standards this stage is held to, and they
live in plugins/compass/shared/ so they are the same for every stage that uses them:
shared/engine.md— how the orchestrator paces a long run, and where it deliberately stops.
(A standard nobody loads is not a standard. shared/MANIFEST declares who reads each file and
doctrine-wired-check.sh proves it — feynman.md sat unread for three releases while its own
first line claimed three stages loaded it.)
Compass is a graph
Compass's lifecycle is an explicit directed graph — drawn at design time, never improvised at runtime:
- The org-graph (fixed): contract → review-contract → plan → review-plan → build → review-build → ship. Nodes are stages; edges are script gates (
compass.sh gate, exit codes) — the next stage is never chosen by a model. - Independent verifier nodes: the adversarial reviews, the cold-critic (fresh subagent, cold screenshots only, 2 consecutive GOs on one tree sha), and the post-ship critique loop (v0.12.0) — a bounded cycle after ship: critique the LIVE system against the contract, loop back on material findings, terminate on N consecutive clean rounds or the cap,
SHIPPEDunwritable untilloop-convergedpasses. A verifier that shares the builder's context shares its blind spots — Compass's critics get fresh context and on-disk evidence only. - Collapsibility rule: every node must justify itself (a distinct specialty, a parallel fan-out, auditable branching, or independent verification) — a node you can't tie to a signal gets collapsed back into the loop it came from.
- Bounded cycles, evidence-based stop rules: every loop carries a cap, a convergence bound read from the contract header, stall/oscillation detection, and budget metering owned by the registration gate itself.
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.
- 2d ago First seen · 114 lines · 93 tokens per session scan A fc992eca9be7
start is a skill published in the GitHub repository Rishi4792/compass (2 stars, last pushed 8d ago), licensed MIT. It adds 93 tokens to every session and 3,831 once invoked, about $0.0005 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.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
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…
agent-host-chat-contributions
Build and review cross-cutting agent-host chat behavior through lifecycle contributions. Use when adding turn lifecycle side effects, prompt or context injection, restored-history transformation, protocol-action observation, or when reviewing changes that add code to AgentSideEffects or AgentService.
auto-perf-optimize
Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.