OpenParable
Skip to content
← Back to skills

Ghx

Productivity
ghx

Ghx is an Agent Skill for productivity published in the majiayu000/claude-skill-registry repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

GitHub enhanced operations skill. Provides unified context collection and publishing commands for PR/issue workflows.

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/majiayu000/claude-skill-registry/blob/HEAD/skills/integration/ghx/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/ghx, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

GHX Skill

ghx is the shared GitHub operations layer used by higher-level skills.

It combines two capabilities:

  • Context collection (issue/PR metadata, comments, diffs, threads, checks)
  • Publishing (PR creation/updates, comments, review replies, inline review posting)

Environment Variables

  • GITHUB_OUTPUT_DIR: artifact directory (caller-provided; defaults to a temporary directory when unset)
  • GITHUB_CONTEXT_DIR: context directory (default ${GITHUB_OUTPUT_DIR}/github-context)
  • GITHUB_TOKEN / GH_TOKEN: GitHub token for gh
  • DRY_RUN: preview mode (true|false)

Commands

Use scripts/ghx.sh as the primary entrypoint.

Contract Boundary

  • Public capability contract (for other skills): call ghx.sh commands and depend on stable output artifacts (for example publish-results.json).
  • Internal implementation contract (within ghx): files like publish-intent.json and action schema details are internal to ghx and may evolve.
  • Other skills should describe desired publish outcomes, not hardcode ghx internal action schemas.

Context

scripts/ghx.sh context collect holon-run/holon#123
scripts/ghx.sh context collect 123 holon-run/holon

Publish from intent (ghx-internal / advanced)

scripts/ghx.sh intent run --intent=${GITHUB_OUTPUT_DIR}/publish-intent.json

Direct publish commands

scripts/ghx.sh review publish --pr=holon-run/holon#123 --body-file=review.md --comments-file=review.json
scripts/ghx.sh pr create --repo=holon-run/holon --title="Title" --body-file=summary.md --head=feature/x --base=main
scripts/ghx.sh pr update --pr=holon-run/holon#123 --title="New title" --body-file=summary.md
scripts/ghx.sh pr comment --pr=holon-run/holon#123 --body-file=summary.md

Output Contract

Context artifacts

Under ${GITHUB_CONTEXT_DIR}:

  • github/pr.json or github/issue.json
  • github/comments.json
  • Optional: github/pr.diff, github/files.json, github/review_threads.json, github/check_runs.json, github/commits.json
  • manifest.json

Publish artifacts

Under ${GITHUB_OUTPUT_DIR}:

  • publish-results.json

Notes

  • ghx is capability-oriented: upper-level skills should focus on final outputs and outcomes.
  • Prefer ghx commands for multi-step or error-prone GitHub operations.
  • Prefer direct capability commands (pr comment, review publish) from external skills.
  • For multi-reply review workflows, use intent run with reply_review actions in publish-intent.json.