agentleFS
jellydnjellydn/my-ai-tools/skills/visual-pr/SKILL.md

visual-pr

Skill127 starsChanged 26 days agoMITView on GitHub

Posts a concise visual outline as a GitHub pull request comment. Use only when the user explicitly asks for visual-pr.

5 other projects use a version of this, 3 of them edited.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. 3 are common, 7 are distinctive, and 4 are very specific to this project.

  • 3

    Common

    Most of the 5 edited versions we read keep these.

  • 7

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use the open pull request for the current branch when one exists.

    line 26
  2. Distinctive:

    If no pull request exists, inspect the branch, commit task changes when needed, push it, and create a draft pull request according to the …

    line 27
  3. Common:

    Ask the user to select a pull request only when the current branch has no relevant work and there is no safe pull request to create.

    line 29
  4. Common:

    Read the ticket and relevant task artifacts.

    line 34
  5. Common:

    Read the complete pull request diff and enough surrounding code to understand behavior and ownership.

    line 35
  6. Very specific:

    Use gh pr view to collect metadata and changed files.

    line 36
  7. Very specific:

    Read $SKILL_PATH/references/comment-template.md, then write the completed comment to a temporary file:

    line 41
  8. Distinctive:

    Keep the outline concise:

    line 47
  9. Distinctive:

    Put only reviewer warnings, migrations, compatibility constraints, deliberate omissions, or surprising decisions in

    line 50
  10. Very specific:

    Special things to note: Use - None. when there are none.

    line 51
  11. Distinctive:

    Use one to three focused views: a data or API shape, pseudocode, a shallow file tree, a component tree, or a call, control, or data flow.

    line 53
  12. Very specific:

    Prefer diff blocks for changes to an existing shape.

    line 55
  13. Distinctive:

    Use real names from the diff and omit categories that did not change.

    line 56
  14. Distinctive:

    Write as one person to another.

    line 76

“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Visual PR

Post a compact visual explanation of a pull request as a comment. Do not replace or append to the pull request description.

Workflow

1. Find or create the pull request

gh pr view --json url,number,title,state,baseRefName,headRefName 2>/dev/null
  • Use the open pull request for the current branch when one exists.
  • If no pull request exists, inspect the branch, commit task changes when needed, push it, and create a draft pull request according to the repository's conventions. Do not use the visual outline as the pull request body.
  • Ask the user to select a pull request only when the current branch has no relevant work and there is no safe pull request to create.

2. Understand the change

  • Read the ticket and relevant task artifacts.
  • Read the complete pull request diff and enough surrounding code to understand behavior and ownership.
  • Use gh pr view to collect metadata and changed files.
  • Identify only the contracts, responsibilities, and runtime paths that a reviewer needs to understand.

3. Write the visual outline

Read $SKILL_PATH/references/comment-template.md, then write the completed comment to a temporary file:

COMMENT_FILE=$(mktemp "${TMPDIR:-/tmp}/visual-pr-comment.XXXXXX")

Keep the outline concise:

  • Explain Why the change in exactly one sentence.
  • Put only reviewer warnings, migrations, compatibility constraints, deliberate omissions, or surprising decisions in Special things to note. Use - None. when there are none.
  • Make Change outline a structural view, not prose or a file-by-file changelog.
  • Use one to three focused views: a data or API shape, pseudocode, a shallow file tree, a component tree, or a call, control, or data flow.
  • Prefer diff blocks for changes to an existing shape. Show the complete target shape when most of it is new.
  • Use real names from the diff and omit categories that did not change.

4. Publish and verify the comment

COMMENT_URL=$("$SKILL_PATH/scripts/publish-comment.sh" <number> "$COMMENT_FILE")
test -n "$COMMENT_URL"
printf '%s\n' "$COMMENT_URL"
rm -f "$COMMENT_FILE"

The script adds a hidden visual-pr marker. It updates the current GitHub user's existing marked comment, or runs gh pr comment when none exists. Thus, one pull request has at most one visual-pr comment from that user. Never use gh pr edit for the visual outline because that replaces the pull request description.

5. Report completion

Reply with the pull request URL, the comment URL returned by GitHub, a two-sentence summary, and the verification that the comment command succeeded.

Write as one person to another. Use simple, concise language and avoid jargon.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.