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/abel30567/fermi-mcp/github-apinpx skills add abel30567/fermi-mcp --skill github-apigit clone --depth 1 https://github.com/abel30567/fermi-mcpWhat 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 | $0.00036 | $0.00735 |
| Opus 5 | $0.00018 | $0.00367 |
| Sonnet 5 | $0.00007 | $0.00147 |
| Haiku 4.5 | $0.00004 | $0.00073 |
Grade A, and why
github-api 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 — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.
GitHub API workflow
Auth
Always call through execute (codemode.fetch_url) or fetch_url with:
- Header
Authorization: Bearer {{secret:GITHUB_TOKEN}}(never inline a raw token; the secret is scoped to api.github.com) - Header
User-Agent: Fermi-Agent(GitHub rejects requests without one) - API base:
https://api.github.com
Common recipes
- Create issue:
POST /repos/{owner}/{repo}/issues{ title, body, labels } - Create PR:
POST /repos/{owner}/{repo}/pulls{ title, head, base, body } - Read file:
GET /repos/{owner}/{repo}/contents/{path}(body is base64) - Write file:
PUT /repos/{owner}/{repo}/contents/{path}{ message, content: <base64>, sha?, branch? }
CRITICAL: UTF-8 encoding (corruption hazard)
atob/btoa alone and Buffer.from(...) in the Worker sandbox corrupt multi-byte
UTF-8 (em-dashes, arrows, smart quotes) — each read-modify-write round trip adds
mojibake. Always use the symmetric pair:
- READ:
decodeURIComponent(escape(atob(b64.replace(/\n/g, '')))) - WRITE:
btoa(unescape(encodeURIComponent(content)))Or the byte-exact TextDecoder/TextEncoder loop for binary-safe handling. Also:→-style escapes only resolve inside JS string literals — when generating JSX/text via templates, insert real unicode characters.
Large binary commits — blob-rename technique (preferred)
The model cannot reliably transcribe >3K chars of base64. When a large file must be committed:
- Have the USER push the file to the repo under any filename.
GET /repos/{o}/{r}/contents/{tmp}to get the blob SHA.POST /repos/{o}/{r}/git/treeswithbase_tree= HEAD tree, adding the blob at the desired path (andsha: nullentries to delete the temp file).POST /repos/{o}/{r}/git/commitswith that tree + parent HEAD.PATCH /repos/{o}/{r}/git/refs/heads/{branch}to the new commit. Zero file bytes pass through model context — only SHA pointers.
Fallback: R2 chunk relay (file not yet on GitHub)
Split base64 into ~1000-char chunks, write each via fs_write to
assets/<name>/<i>, then one execute reads all chunks, concatenates, verifies
magic bytes + length, and PUTs to GitHub with the auth header above.
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 · 51 lines · 36 tokens per session scan A 0145f9b19680
github-api is a skill published in the GitHub repository abel30567/fermi-mcp (1 stars, last pushed 5d ago), licensed MIT. It adds 36 tokens to every session and 735 once invoked, about $0.0002 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
issue
Use when starting a chain from a GitHub issue — turning an issue URL or number into a triaged, planned, dispatched, and reviewed pull request. Classifies the thread (bug → root-cause discipline, feature → plan chain, question → drafted reply), synthesizes a spec from the issue's own acceptance criteria, then runs the…
github
Prepare GitHub issues, pull requests, reviews, and release notes for Zhin projects. Use when asked to write an issue, PR description, changelog entry, or review comment. Triggers: 写 issue, PR 描述, 提 PR, release notes, gh pr.
github-integration
GitHub 集成 Skill,支持创建/查询 Issue、PR,获取仓库信息。用于 Agent 与 GitHub 协作。.
horse-database-pooling
Guide for setting up thread-safe database connection pooling (FireDAC / UniDAC) in multithreaded Horse applications.
horse-mvc-architecture
Guide for structuring corporate Horse applications using Clean MVC (Model-View-Controller) principles and decoupling HTTP layers from business logic.
horse-request-response
Guide to interacting with THorseRequest (body, query, params, headers) and THorseResponse (Send, Status, ContentType).