Instructions for publishing approved local code changes as a monitored GitHub draft pull request. A draft pull request is a reviewable change that is not yet marked ready to merge.

Use it when you have authorized changes that need a branch, commit, push, draft pull request, and follow-up on CI checks.

It sets checks for scope, repository rules, testing, staging, and review before changes are pushed and shared.

pr-draft — Skill for Claude Code by yarlson/yarstack

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.

Any agent

npx skills add yarlson/yarstack --skill pr-draft

Clone the repo

git clone --depth 1 https://github.com/yarlson/yarstack

Made for: Claude Code.

Or install yarstack, the plugin that ships this one along with the rest of its 46 skills, 1 agent.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/yarlson/yarstack/pr-draft/github.svg)](https://agentmods.dev/skills/yarlson/yarstack/pr-draft)

Your own site

<a href="https://agentmods.dev/skills/yarlson/yarstack/pr-draft"><img src="https://agentmods.dev/badge/skills/yarlson/yarstack/pr-draft/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/yarlson/yarstack/pr-draft"><img src="https://agentmods.dev/badge/skills/yarlson/yarstack/pr-draft.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 43 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 560 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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.

Model	Per session	Once invoked
Fable 5.1	$0.00043	$0.00560
Opus 5.5	$0.00017	$0.00224
Sonnet 5.5	$0.00009	$0.00112
Haiku 4.5	$0.00004	$0.00056

Measured 3d ago against content hash f7e60616f26b, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

pr-draft 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 3d 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.

plugins/yarstack/skills/pr-draft/SKILL.md · 27 lines

How it starts

The opening of the file, as written. The whole thing — 27 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Draft PR

Own end-to-end draft-PR publication after implementation, validation, and review are complete.

Workflow

Confirm shipping authority, repository rules, remotes, current branch, exact scope, and completed validation and review evidence. Resolve the base branch from live remote evidence, such as git ls-remote --symref <remote> HEAD, unless the user names it.
Preserve unrelated staged, unstaged, and untracked work. Stop only when safe isolation is impossible or a scoped path is sensitive or unexpectedly generated.
Create a short branch when needed, stage explicit paths with git add -- <paths>, and review the staged path list and diff.
Run required checks when existing evidence does not cover the exact staged content.
Follow repository PR templates, title conventions, and required fields. Otherwise write concise evidence-based Problem, Change, Tests, and only necessary Notes sections.
Use text-improve on the completed title and body without changing verified facts, required template wording, or technical terms.
Use change-cleanup-review on the complete staged change and proposed PR presentation. Stop and report material cleanup findings unless the parent request separately authorizes repair. After an authorized repair, restage the exact scope and rerun invalidated checks and prose improvement.
Commit using repository conventions and push with an upstream when needed.
Check whether a pull request already exists for the branch with gh pr view --json url,isDraft; if one exists, update it instead of opening another. Otherwise run gh pr create --draft --base <base> --title <title> --body-file <file>. Monitor the checks for the pushed commit with gh pr checks <number> --watch --fail-fast under the host's command timeout.
Diagnose failures from logs. Route behavior changes through behavior-implement and test-design, and specialist failures through their owning workflow; verify, rerun change-cleanup-review when corrections materially move the change, then commit and push only scoped corrections.
Stop when checks pass, the user permits cancellation, a concrete external blocker prevents progress, or three fix cycles for the same failing check have not fixed it.

Read the full file on GitHub · 27 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.

3d ago Changed f7e60616f26b
1mo ago First seen · 27 lines · 43 tokens per session scan A ed6734c20fda

Subscribe to this mod's changes

pr-draft is a skill published in the GitHub repository yarlson/yarstack (3 stars, last pushed 2d ago), licensed MIT. It adds 43 tokens to every session and 560 once invoked, about $0.0002 per session on Opus 5.5. 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.

Code review Git and workflow agent-skills agent-workflows ai-coding coding-agents developer-tools

owndiff-pre-commit

Audit OwnDiff changes before committing or releasing by checking scope, generated artifacts, private data, documentation accuracy, version consistency, structured files, tests, and the OwnDiff ownership gate. Use when a user asks to verify, clean up, prepare, commit, publish, or release changes in the OwnDiff…

owndiff/own-your-diff · 77 tokens

git-review

Review a repository diff for accidental changes, secrets, generated files, dependency and migration impact, tests, documentation, and scope before completion.

Arnosdeus/agent-toolkit · 30 tokens

gentle-ai-collab-perfect

Trigger: contributing to Gentleman-Programming/gentle-ai as an external collaborator. Strict issue-first workflow, honest PR bodies, contributor-vs-maintainer scope, chained-PR strategy, verification protocol, docstring coverage. Load whenever the active repo is Gentleman-Programming/gentle-ai and any part of the…

Gentleman-Programming/gentle-ai · 106 tokens

pr-draft