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/cozystack/ccp/debugnpx skills add cozystack/ccp --skill debuggit clone --depth 1 https://github.com/cozystack/ccpWhat 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 | $0.00157 | $0.03708 |
| Opus 5 | $0.00078 | $0.01854 |
| Sonnet 5 | $0.00031 | $0.00742 |
| Haiku 4.5 | $0.00016 | $0.00371 |
Grade A, and why
debug 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 — 239 lines — stays where its author put it; the contents beside it link to each section on GitHub.
cozystack:debug
Work in reasoning mode. Use the phrasing cozystack:debug. Announce phase transitions: cozystack:debug Phase N — <name>.
Note on language in this SKILL.md — every operator-facing prompt below is written in English for clarity. At runtime the skill matches the operator's natural language detected from prior conversation messages (or read from
<config-dir>/.state.yamloperator_languagewhen the wizard chain is in progress). Code identifiers, commands, file paths, and any text destined for GitHub stay canonical.
What this skill does, in one sentence
When something in a Cozystack install is broken or stuck, find out what, classify whose fault, fix what's fixable locally, and on approval prepare a clean upstream issue for what isn't.
Two invocation paths
- From
cozystack:wizard— automatically dispatched when any chain step writesstatus.<skill>.failed_atto<config-dir>/.state.yaml. Inheritsconfig_dir,cluster.context, and the failing skill's error string. - Direct — operator invokes
/cozystack:debugagainst an already-running cluster. Interviews for--config-dir(default$PWD),--context(defaultkubectl config current-context), and what hurts.
Core principles
- Match the operator's natural language. Read from
state.operator_language(set bycozystack:wizardPhase 0) or, when invoked directly, detect from the operator's previous messages in the conversation. Use that language in every prompt, AskUserQuestion option, summary, and gate. Never ask "what language?" separately. Code identifiers, command examples, file paths, and any text that ends up in a GitHub issue body stay in their canonical form (usually English). - Read first, mutate second. Phases 1–4 are read-only.
- One valid path → just do it. When Phase 3 classification + Phase 4 action are unambiguous (operator error with a documented fix, config drift with a clear restore command), the skill applies the fix without an extra "ok to apply?" question — the symptom + classification + proposed action were already shown. Gates remain for (a) destructive workarounds (resource recreation, secret regeneration), (b) issue / PR drafts that go public (Phase 5 explicit yes/no per filing).
- Front-load the interview. The only question the debug skill should ask is in Phase 4 when classification + proposed action are ready: a single screen with symptom, classification, root cause, proposed fix / workaround, and the issue-filing question (yes / no / show-draft-first). The operator either approves the lot or names what to change. Phases 1–3 (symptom gathering, doc check, classification) are read-only and run before that screen.
- Layer-pure operator output. The skill never says "returning control to wizard", "the wizard will retry the failing skill", or any other orchestration commentary in the operator-facing summary. Whoever invoked the skill (a human directly, or the wizard's auto-dispatch on
failed_at) figures out what's next on their own. Internal SKILL.md references tocozystack:wizardare fine for documentation;wizarddoes not appear in any text shown to the operator. - Classify before acting. The same symptom can be operator error or upstream bug; the fix is different.
- Doc-check is not optional. Cozystack documentation is the operator's contract; verify the operator's setup against the relevant page before declaring upstream fault.
- One filing at a time. If multiple issues surface, finish one before starting the next — operators get confused fast.
- No PRs. v1 only drafts issue bodies. Patches that the operator wants to upstream are out of scope and stay manual.
- No silent filings. Every issue draft is shown for approval. If the operator doesn't have a GitHub account, no problem — the diagnostic bundle is still useful to them.
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 · 239 lines · 157 tokens per session scan A 1868a2262f06
debug is a skill published in the GitHub repository cozystack/ccp (4 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 157 tokens to every session and 3,708 once invoked, about $0.0008 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…