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/cairn-framework/cairn/cairn-loop-implementnpx skills add cairn-framework/cairn --skill cairn-loop-implementgit clone --depth 1 https://github.com/cairn-framework/cairnWhat 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 | $0.00065 | $0.00778 |
| Opus 5 | $0.00032 | $0.00389 |
| Sonnet 5 | $0.00013 | $0.00156 |
| Haiku 4.5 | $0.00006 | $0.00078 |
Grade A, and why
cairn-loop-implement 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 — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Implement one loop unit
Loaded by cairn-dev loop mode at its Implement and test step. Inputs: the
selected unit, its resolved node, the success criterion from cairn-loop-scope,
and the bound $CAIRN.
Declared exits, exactly one, as the last line you return to loop mode:
IMPLEMENTED: the change is written and its test exists; loop mode runs Verify.LOOP HALTED: implementation cannot proceed without a maintainer.
1. Get on the right branch
The branch is loop/<tail>, where <tail> is the derived form from loop
mode's Isolation rule: todo.<slug>, <finding-code>.<node>, or
split.<slug>. Every later step (push, PR, Cleanup) uses this exact name.
If it is already checked out (adopted at verdict time, or created earlier this session during MISSION materialisation or decomposition), continue on it. Otherwise create it from fresh origin/main:
git checkout --detach origin/main && git checkout -b loop/<tail>
If the derived name already exists but was NOT adopted by the verdict and NOT created this session, you missed a preflight row. Do not improvise and do not halt here: return to loop mode's preflight table, which owns that state and can adopt or quarantine the branch. Say which name collided.
2. Make the smallest change that satisfies the criterion
Not the best change you can think of. The smallest one that makes the criterion true. Anything else is a separate unit.
- Touch only what the criterion requires. Do not improve adjacent code, reformat, or refactor things that are not broken. Match the surrounding style even where you would have chosen differently.
- Remove imports, variables, and helpers that YOUR change orphaned. Leave pre-existing dead code alone; mention it instead.
- If you write far more code than the criterion needs, stop and reconsider. A senior engineer calling the result overcomplicated is a defect.
3. Keep the blueprint honest
The graph must still describe the code when you are done:
- Every new file falls under a node
path, tests included. If none fits, extend a node's paths or declare a new Module. - Every new cross-module call gets a blueprint edge. Check for a cycle first:
$CAIRN deps <target> --transitive. - User-facing strings go to the copy location named by loop mode's Repo bindings. Never hardcode them in source, and never guess the path: read the binding.
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 · 77 lines · 65 tokens per session scan A 494752565288
cairn-loop-implement is a skill published in the GitHub repository cairn-framework/cairn (40 stars, last pushed 3d ago), licensed Apache-2.0. It adds 65 tokens to every session and 778 once invoked, about $0.0003 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 2026-08-30.
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.
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.
chat-perf
Run chat perf benchmarks and memory leak checks against the local dev build or any published VS Code version. Use when investigating chat rendering regressions, validating perf-sensitive changes to chat UI, or checking for memory leaks in the chat response pipeline.
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…