Github logo

Github

Organization
vanillagreencom
github

Load to work a GitHub pull request: threads, comments, reviews, CI logs, merges.

Overview

Publishervanillagreencom
Repositorykendex
Skill namegithub
Stars
80
Forks
31
Bundled files
61
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 61 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by vanillagreencom on GitHub. Read the source before you install it.

Installation

Install the Github AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/vanillagreencom/kendex.git /tmp/kendex
mkdir -p .claude/skills
cp -r /tmp/kendex/skills/github .claude/skills/github
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Github in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Github on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Github is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

GitHub Queries

bash
.agents/skills/github/scripts/github.sh [-C <path>] <command> [options]

Commands

CommandPurpose
pr-data <N> [--actionable]Get PR with threads, comments, files. --actionable: unresolved non-outdated only.
pr-view [N] [--json FIELDS]View PR details (wraps gh pr view with bounded auth/no-PR errors)
pr-threads <N> [--unresolved|--resolved] [--format=safe|raw]Complete paginated thread list/count, outdated included. Both filters apply in both formats. See PR blocked with no visible conversations.
pr-list-ready [--all] [--format=safe|table]List PRs ready for merge
pr-list-failing [--all] [--format=safe|table]List PRs with CI failures
pr-create [--title T] [--body B | --body-file PATH] [--draft] [--dry-run] [--force]Create PR as bot. Safety checks: not main, has commits, pushed; --force skips them.
pr-edit-body <N> --body-file PATHUpdate an existing PR body through the sanitized router.
pr-merge <N> [--check|--force|--admin|--auto]Merge PR. --check reports readiness as JSON on stdout plus a one-word verdict and head-run: <ids> (the run scope of the CI classification) on stderr; --auto queues a currently-blocked PR. Three exit codes, the review-thread gate, and --force/--admin. See PR Merge Outcomes.
ci-classify-refusal <N>Name the cause of a pr-merge refusal on one cause: line (fetch_error, merge_conflict, changes_requested, threads, ci_failed, ci_pending, computing, merged, closed, none; an issue prefix outside that vocabulary becomes the cause word itself, and none means the checks pass now); ci_failed adds fail: lines run-correlated to the authoritative run and superseded: lines naming runs whose checks were not counted. --help
pr-cross-check [N...] [--quick|--verify]Cross-PR analysis. --verify: full build+test (auto-detects build system).
pr-issue <N> [--format=safe|text]Extract issue ID from PR branch (configurable via GH_ISSUE_PATTERN)
label-add <PR-or-issue> <label> [--issue] [--required|--optional]Add a label after checking the live inventory. Mode semantics and exit codes: label-add --help.
label-remove <PR-or-issue> <label> [--issue]Remove a label through the sanitized router.
await-mergeable <N> [--interval S] [--max-iter N] [--quiet]Block until GitHub resolves a PR's merge state. Polls state + mergeStateStatus. Exit 0 + JSON on resolve, 124 on timeout.
ci-logs <N> [--lines N] [--format=safe|text]Get CI failure logs for PR
bot-token [--format=safe|text]Check if bot token is configured, naming the selected variable as source
dismiss-review <PR> [--bot|--user NAME] [--message M]Dismiss blocking review
resolve-thread <PRRT_...>Mark thread(s) resolved. Works on threads the UI cannot render. See PR blocked with no visible conversations.
unresolve-thread <PRRT_...>Reopen thread(s)
post-reply <PRRT_...|numeric-id> [body | --body-file PATH] [--pr N]Reply to review comment. --pr N is REQUIRED for numeric comment IDs; thread PRRT_... IDs need no PR number.
post-comment <PR> [body | --body-file PATH]Post PR-level comment.
find-comment <PR> --pattern <regex>Find comment by pattern/author
edit-comment <id> [body | --body-file PATH]Edit existing comment.
sticky-comment <PR> [--verdict|--analysis|--body]Get bot sticky comment. --verdict: quick pass/fail. --analysis: deep recommendation.

CI waiting belongs to .agents/skills/orch/scripts/ci-wait; await-mergeable waits for merge-state resolution.

Contracts: label-add --help, git-https-auth --help, git-diff-summary --help.

PR Merge Outcomes

Full contract: pr-merge --help. Exit 75 is volatile: the caller arms one exact head and waits on that head with the orch skill's queue-wait, whose --help § Verdicts maps each verdict to a route; an unrecognized verdict is never re-armed. With the review-gate skill installed, its watcher output contract is pr-watch.sh --help. If can_merge is false with no issues, read state. The thread gate is Policy, not mechanism. --force and the explicit-user-only --admin are its overrides.

PR blocked with no visible conversations

Under required_conversation_resolution, an outdated thread can block a merge while the UI shows none; resolve-thread reaches it by id.

bash
github.sh pr-threads 42                  # complete list, outdated included
github.sh resolve-thread PRRT_kwDO...    # resolve by thread id

pr-threads follows every page and fails rather than returning a partial list, so a thread id absent from its output is genuinely absent. Repeat resolve-thread per blocking id until the merge clears.

Waiting for merge state

Never gate termination on gh pr view --json mergeable. That field stays UNKNOWN permanently after a merge. Use await-mergeable (resolution rules and exit codes: await-mergeable --help). To watch MANY PRs, do not hand-roll a poll loop keyed on state transitions. Use the review-gate skill's reducer when installed (.agents/skills/review-gate/scripts/pr-watch.sh).

Output Formats

Formats and flag rules: github.sh --help.

Configuration

Keep secrets in .env.local; commit non-secret defaults to kendex.settings.toml under [env]. Other contracts: github.sh --help.

Troubleshooting

VAR_SIGN: use a multi-line GraphQL query with -F variables.

Stale-token HTTP 401: clear both environment tokens:

bash
env -u GH_TOKEN -u GITHUB_TOKEN gh pr list

github.sh falls back when keyring auth succeeds.

Dependencies

  • gh CLI (authenticated)
  • jq
  • op CLI (optional, 1Password token references)

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

and 1 more files.

Frequently asked questions

What does the Github AI skill do?

Load to work a GitHub pull request: threads, comments, reviews, CI logs, merges.

Why use Github on TypingMind?

Because you install it once and use it with any model. Github is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Github in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/vanillagreencom/kendex/tree/main/skills/github. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Github?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Github?

As many as you like. As long as a model supports skills, you can use Github with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Github AI skill free?

Yes. It is published on GitHub by vanillagreencom under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇