github

A connection to GitHub through the already signed-in `gh` command-line tool. It can work with repositories, issues, pull requests, code searches, reviews, diffs, automated checks, and GitHub's REST API.

In plain words
What is it for?
Use it to find repositories, issues, pull requests, commits, or code; inspect changes and review discussions; check failed automated tests; and call GitHub's REST API.
Why use it?
It avoids writing a separate GitHub client or asking for an access token. It also helps identify account, permission, or repository problems when a request fails.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/sybil-solutions/local-studio/github
Any agent
npx skills add sybil-solutions/local-studio --skill github
Clone the repo
git clone --depth 1 https://github.com/sybil-solutions/local-studio

Made for: Claude Code, Codex.

Per session 67 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 850 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00067 $0.00850
Opus 5 $0.00034 $0.00425
Sonnet 5 $0.00013 $0.00170
Haiku 4.5 $0.00007 $0.00085

Measured 2d ago against content hash 52d1d1281293, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

github 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.

frontend/desktop/resources/skills/github/SKILL.md · 36 lines

How it starts

The opening of the file, as written. The whole thing — 36 lines — stays where its author put it; the contents beside it link to each section on GitHub.

github

The github_* tools are the gh CLI, already installed and already authenticated on this machine. That is the whole design: gh holds the credentials in the OS keyring, knows which repo the session directory belongs to, paginates, and renders diffs and check runs. Do not write an API client, do not ask the user for a token, and do not go looking for one in the environment.

github_status tells you which account you are acting as and which repository this directory resolves to. Call it when something 404s or 403s — the cause is almost always the wrong account, a missing token scope, or a directory that is not a GitHub checkout, not a bad query.

Tools

  • github_status — signed-in account, scopes, and the repo this directory resolves to.
  • github_search — issues, PRs, repos, code, or commits across GitHub, using GitHub search syntax.
  • github_issue_list / github_issue_view — one repo's issues; the view includes the comment thread.
  • github_pr_list / github_pr_view — pull requests; the view includes reviews and comments.
  • github_pr_diff — the diff, or with namesOnly just the changed paths.
  • github_pr_checks — CI check runs for a PR. A non-zero exit is the answer, not an error.
  • github_run_list / github_run_view — Actions runs; failedLogs returns only the failing steps' output.
  • github_api — any REST endpoint the named tools miss. GET only.
  • github_cli — any other gh subcommand, as an argv array. This is the write path.

Every tool takes an optional repo as OWNER/REPO. Omit it to use the repository checked out in this session's project directory, exactly as running gh there would.

Protocol

  1. Narrow before you read. github_pr_diff with namesOnly, or github_issue_list before github_issue_view, costs one call and saves a truncated wall of output.
  2. Read the comments. A PR's blocking objection and an issue's actual decision live in the thread, not the description — github_pr_view and github_issue_view include them by default, so use them rather than summarizing from the title.
  3. For red CI, go github_pr_checksgithub_run_view with failedLogs. That is the shortest path from "the build broke" to the error line, and it is far smaller than a whole run's logs.
  4. Prefer a named tool over github_cli when one exists; the named tools return structured JSON, github_cli returns whatever gh prints.
  5. github_cli writes. gh pr create, gh issue comment, gh pr merge, gh release create are public, attributed to the user, and usually permanent. Only run one the user actually asked for, and say exactly what you ran.
  6. There is no shell behind github_cli — pass each argument as its own array element. Pipes, redirects, globs, and && do not work.
  7. Credential subcommands (gh auth, gh secret, gh config) and every delete are refused by design. If the user genuinely wants one, tell them to run it themselves rather than working around the block.
  8. For a repository that is checked out locally, plain git and reading the files is faster and more accurate than GitHub search. Use these tools for what only GitHub knows: issues, reviews, CI, and other people's repos.

Read the full file on GitHub · 36 lines

Changes

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.

  1. 2d ago First seen · 36 lines · 67 tokens per session scan A 52d1d1281293

Subscribe to this mod's changes

github is a skill published in the GitHub repository sybil-solutions/local-studio (1,745 stars, last pushed 5d ago), licensed Apache-2.0. It adds 67 tokens to every session and 850 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.