pull-request-msg-with-gh
Skill19 starsChanged 8 months agoMITView on GitHub
Generate a PR_MESSAGE.md file from session context using GitHub CLI. Detects related issues via branch-keyword search, writes a structured PR description with commit subject, summary, test plan, and changelog. Validates with commitlint and markdownlint. Use when preparing a pull request on GitHub.
- Installs packages
In this repository53 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .claude/skills/agent-rules/AGENTS.mdAGENTS.md
- .github/workflows/AGENTS.mdAGENTS.md
- pkg/AGENTS.mdAGENTS.md
- agent-ready-goSkill
- agent-rulesSkill
- architecture-designerSkill
- cli-developerSkill
- cloudformation-to-pulumiSkill
- code-reviewerSkill
- commitlintSkill
- commitSkill
- csharp-developerSkill
- decideSkill
- dep-upgradeSkill
- design-principlesSkill
- devops-engineerSkill
- flux-cliSkill
- flux-operator-cliSkill
- golang-proSkill
- go-nolint-auditSkill
- javascript-proSkill
- kubernetes-specialistSkill
- lint-fixSkill
- markdownlintSkill
- prSkill
- pulumi-arm-to-pulumiSkill
- pulumi-automation-apiSkill
- pulumi-best-practicesSkill
- pulumi-cdk-to-pulumiSkill
- pulumi-cliSkill
- pulumi-componentSkill
- pulumi-escSkill
- pulumi-goSkill
- pulumi-neoSkill
- pulumi-pythonSkill
- pulumi-terraform-to-pulumiSkill
- pulumi-typescriptSkill
- python-proSkill
- reviewSkill
- roadmapSkill
- rust-async-patternsSkill
- rust-best-practicesSkill
- rust-engineerSkill
- scoutSkill
- security-auditSkill
- spring-boot-engineerSkill
- sre-engineerSkill
- sync-websiteSkill
- tailscale-installSkill
- tech-debtSkill
- the-foolSkill
- typescript-proSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 1 is distinctive, and 6 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 26
If gh is not installed, stop and report:
- Very specific: line 30
After install, run gh auth login to authenticate
- Very specific: line 32
Do not proceed without a working, authenticated gh CLI.
- Very specific: line 41
If the user indicates there is no associated issue, omit the Closes #<issue-number> footer entirely.
- Very specific: line 97
If the commitlint and markdownlint skills are available in the current context, use them to validate PR_MESSAGE.md.
- Very specific: line 110
If neither the skills nor npx are available, fail hard:
- Distinctive: line 116
Do not silently skip validation.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
<!-- Copyright 2025-2026 Richard Shade. Licensed under Apache-2.0. --><!-- SPDX-License-Identifier: Apache-2.0 -->Pull Request Message Generator
Generate a PR_MESSAGE.md file summarizing the current session's work.
Requires gh CLI authenticated with the target repository.
Prerequisite check
gh --version 2>/dev/null && gh auth status 2>/dev/nullIf gh is not installed, stop and report:
- Install from https://cli.github.com
ghcannot be installed via npm- After install, run
gh auth loginto authenticate
Do not proceed without a working, authenticated gh CLI.
Workflow
- Analyze session context — review changes made in the current
session. Explore changed files with
git diffandgit statusto understand the full scope of work, in addition to reviewing conversation history. Both sources inform the PR message. - Detect issue number — use the issue detection algorithm below.
If the user indicates there is no associated issue, omit the
Closes #<issue-number>footer entirely. - Determine commit type — based on the nature of changes (see
references/pr-message-format.mdfor the list of types) - Write PR_MESSAGE.md — using the format in
references/pr-message-format.md - Add to .gitignore — add
PR_MESSAGE.mdif not already present - Validate — run commitlint and markdownlint checks
Issue number detection
The issue number MUST be determined by searching GitHub, NOT by extracting numbers from branch names.
Algorithm
# Step 1: Get branch name
BRANCH=$(git branch --show-current)
# Step 2: Extract keywords (strip leading numbers and hyphens)
# Example: "125-greenops-equivalencies" → "greenops equivalencies"
KEYWORDS=$(echo "$BRANCH" | sed 's/^[0-9]*-//' | tr '-' ' ')
# Step 3: Search GitHub issues
gh issue list --search "$KEYWORDS" --state all \
--json number,title --limit 5Evaluation rules
- Exactly 1 result with title containing most keywords → use it
- 0 results → ask the user for the issue number
- 2+ results → show options and ask the user to pick
- Ambiguous match → ask the user to confirm
Never extract issue numbers from
- Branch name digits (e.g.,
125-is NOT issue #125) - Spec folder names (e.g.,
specs/125-xyz/) - Worktree folder names
These are internal identifiers, not GitHub issue numbers.
Critical rules
- No redundancy — summary must add context beyond the commit subject
- No
## Commit Messagesection — the first line IS the commit message - Never modify
.markdownlint.json— use CLI flags instead - Trailing newline — file must end with a single newline
- Ask when ambiguous — if issue search returns 0, 2+, or unclear results, ask the user
Validation
If the commitlint and markdownlint skills are available in the current context, use them to validate PR_MESSAGE.md. Otherwise, run directly:
# Validate commit message format
cat PR_MESSAGE.md | npx commitlint
# Validate markdown (MD041 disabled — first line is commit subject,
# not heading)
npx markdownlint --disable MD041 -- PR_MESSAGE.mdIf neither the skills nor npx are available, fail hard:
- Install commitlint:
npm install -g @commitlint/cli @commitlint/config-conventional - Install markdownlint:
npm install -g markdownlint-cli
Do not silently skip validation.
Fix any issues and re-validate before reporting success.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

