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 skills add rjwalters/kicad-tools --skill loom-ci-refusals-referencegit clone --depth 1 https://github.com/rjwalters/kicad-toolsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/rjwalters/kicad-tools/loom-ci-refusals-reference)<a href="https://agentmods.dev/skills/rjwalters/kicad-tools/loom-ci-refusals-reference"><img src="https://agentmods.dev/badge/skills/rjwalters/kicad-tools/loom-ci-refusals-reference/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/rjwalters/kicad-tools/loom-ci-refusals-reference"><img src="https://agentmods.dev/badge/skills/rjwalters/kicad-tools/loom-ci-refusals-reference.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What 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 5.1 | $0.00073 | $0.01002 |
| Opus 5.5 | $0.00029 | $0.00401 |
| Sonnet 5.5 | $0.00015 | $0.00200 |
| Haiku 4.5 | $0.00007 | $0.00100 |
Grade A, and why
loom-ci-refusals-reference 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 1mo 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 — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CI Refusals: Permission vs Rate Limit, and Re-runs (#10633)
Load when: loom-daemon forge wait-checks printed LOOM-CHECKS-ERROR naming
an HTTP 401/403/429, it left a stderr note about legacy commit statuses,
or a job needs a re-run (cancelled by a runner shutdown, or flaky).
Why this exists
GitHub uses HTTP 403 both for "this credential may not do that" and for
"slow down". Before #10633 a Doctor saw only HTTP 403. It could not tell a
missing GitHub App permission from a secondary rate limit, so it handed PRs
back with CI unverified. When it could not re-run a cancelled job, it rebased
and pushed instead, which cost a full CI cycle and opened a new conflict window.
Reading a wait-checks refusal
wait-checks classifies every refused read, and the class is in the reason:
| Reason contains | Meaning | What to do |
|---|---|---|
permission (needs <perm>) |
The credential lacks that App permission. | Waiting will not help. Quote the line in your hand-back comment. The operator grants <perm> on the App. |
secondary-rate-limit / rate-limit (after read-failed:) |
GitHub throttled the reads for the whole wait. | Say CI is unverified because of a rate limit, not a permission problem. Judge re-checks later. |
credential |
The token was refused (401). |
Report it. Do not retry. |
forbidden |
GitHub refused the action itself. Its message follows. | Read the message. |
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.
- 1mo ago First seen · 68 lines · 73 tokens per session scan A 314cd8a4bc17
loom-ci-refusals-reference is a skill published in the GitHub repository rjwalters/kicad-tools (73 stars, last pushed yesterday), licensed MIT. It adds 73 tokens to every session and 1,002 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-10-08.
loom-ci-refusals-reference compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| loom-ci-refusals-referencerjwalters/kicad-tools | Skill | 73 | yesterday | MIT | A | 73 | Claude Code, Codex |
| baby-sitlangchain-ai/open-swe | Skill | 11k | today | MIT | A | 30 | Claude Code, Codex |
| chinese-git-workflowjnMetaCode/superpowers-zh | Skill | 8.5k | 14d ago | MIT | A | 69 | Claude Code |
| atmos-hookscloudposse/atmos | Skill | 1.4k | today | Apache-2.0 | A | 46 | Claude Code |
| atmos-procloudposse/atmos | Skill | 1.4k | today | Apache-2.0 | A | 38 | Claude Code |
| start-temps-clustergotempsh/temps | Skill | 904 | yesterday | Apache-2.0 | A | 204 | Claude Code, Codex |
| pr-watchSethGammon/Citadel | Skill | 927 | 4d ago | MIT | A | 46 | Claude Code |