From dbud
Use when posting validated code review findings as GitHub PR line-level comments or a GitHub PR review after completing a review.
How this skill is triggered — by the user, by Claude, or both
Slash command
/dbud:add-pr-commentsThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Post validated code review findings from the current review as a GitHub PR review with line-level comments. Use this only after findings have been reviewed, deduplicated, and validated against the current diff.
Post validated code review findings from the current review as a GitHub PR review with line-level comments. Use this only after findings have been reviewed, deduplicated, and validated against the current diff.
Resolve the PR.
gh pr view --json number,headRefOid,url -q '{number: .number, sha: .headRefOid, url: .url}'Confirm review language.
Use the final finding list.
file:line or file:startLine-endLine), why it matters, and recommended fix.Classify findings.
file:line or file:startLine-endLine format.Filter and deduplicate.
low confidence findings unless severity is high or critical.Build the GitHub review payload.
gh repo view --json nameWithOwner -q .nameWithOwner/repos/{owner}/{repo}/pulls/{pr_number}/reviews.commit_id: HEAD commit SHA from step 1event: APPROVE for ready to approve, otherwise REQUEST_CHANGESbody: one-sentence TL;DR of the PR's overall healthcomments: line-anchored comments, each with:
path: relative file path from repo rootline: line number, or range end linestart_line: range start line, only for multi-line rangesside: RIGHTbody: formatted finding commentFormat each comment body.
**[SEVERITY] Title**
Why it matters.
**Recommended fix:** actionable fix
Use severity values CRITICAL, HIGH, MEDIUM, or LOW.
Post the review.
gh api ... --input <payload-file>.Handle errors.
422: move comments that cannot be anchored to the diff into the review body, then retry.403 or 404: report that the gh token or user may lack write access.Report the result.
APPROVE or REQUEST_CHANGES.Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub tkxkd0159/agent-plugins --plugin dbud