Draft Pull Request

What it does

When your agent uses it

• Tasks that involve Pull requests

Example prompts

• “/draft-pull-request”

Requirements

• Compatibility (from SKILL.md): cline, claude, opencode, amp, codex, gemini, cursor, pi

Workflow steps

5 steps, taken from the step headings in SKILL.md.

1. 1Preflight: find or prepare the PR
2. 2Gather only the context you need
3. 3Write the description
4. 4Publish
5. 5Confirm and report

What it can do on your machine

Read from SKILL.md and the folder at commit 163951e. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Shell commands in SKILL.md call:

  • gh
  • git

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md. Its commands use gh and git, which can reach the network depending on how they are called.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

• Compatibility

  cline, claude, opencode, amp, codex, gemini, cursor, pi

  From compatibility in the SKILL.md frontmatter.

Context cost

Draft Pull Request loads about 1.4k tokens when it runs, and up to ~2.7k if it reads all its reference files. Until then it costs about 20 tokens; SKILL.md has 466 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~20

When it runs · the whole SKILL.md, loaded when a task matches

~1.4k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~2.7k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from jellydn/my-ai-tools at commit 163951e, republished under its MIT licence (© jellydn). 466 words, ~1,441 tokens.

name

draft-pull-request

description

Draft pull requests with structured descriptions using gh CLI

compatibility

cline, claude, opencode, amp, codex, gemini, cursor, pi

license

MIT

hint

Use when creating or updating a draft pull request with a structured description using gh CLI

user-invocable

true

disable-model-invocation

true

metadata.audience

all

metadata.workflow

git

/draft-pull-request [title]

BASE_BRANCH=$(gh repo view --json defaultBranchRef -q '.defaultBranchRef.name')
git branch --show-current
git status --short --branch

# Reuse an existing PR for this branch instead of opening a duplicate
gh pr view --json number,url,title,state,isDraft,baseRefName,body 2>/dev/null

git fetch origin "$BASE_BRANCH" --quiet
git log "origin/$BASE_BRANCH"..HEAD --oneline
git diff "origin/$BASE_BRANCH"...HEAD --stat
git diff "origin/$BASE_BRANCH"...HEAD

BODY_FILE=$(mktemp "${TMPDIR:-/tmp}/pr-body.XXXXXX")

# Never force-push.
git push -u origin HEAD

# New PR
gh pr create --draft --base "$BASE_BRANCH" --title "<PR title>" --body-file "$BODY_FILE"

# Existing PR from step 1
gh pr edit <number> --body-file "$BODY_FILE"

gh pr view --json url,title,isDraft -q '"\(.url) draft=\(.isDraft) \(.title)"'
rm -f "$BODY_FILE"

- PR: [#<number> <title>](<url>) (<draft or ready for review, from isDraft>)
- Summary: <2-3 sentences: what the PR does and the key decision>
- Validation: <commands run and their results, or "not run">
- Left out: <uncommitted or unrelated changes, or "none">

Closes #142

## What

Add a refresh-token endpoint and automatic token renewal in the API client.

## Why

Users are logged out when the 1-hour access token expires, which interrupts long sessions.

## How

- `POST /auth/refresh` validates the HttpOnly refresh cookie and issues a new access token.
- The API client retries a `401` response once, after a refresh, before it shows an error.

```diff
 request(url)
   send with access token
+  if status is 401 and not retried
+    refreshToken()
+    retry once
   return response
```

## Reviewer notes

- Refresh tokens rotate on each use; a reused token revokes the whole session.

## Validation

- `bun test src/auth` — 24 passed
- Manual: session stays signed in after the access token expires

© jellydn, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 2 other files (references) in skills/draft-pull-request of jellydn/my-ai-tools.

• SKILL.md
• references/change-outline.md
• references/pr-body-template.md

Open the folder on GitHubat commit 163951e