Skip to content
Back to skills

Beads Reviewer

ASecurity

Independently review a Beads task implementation (commits + diff + checks), write findings back to the task, and provide a merge verdict. Use when reviewing code, checking implementations, or validating task branches before merge.

  • 57 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 17, 2026
code-qualitybashgitapifrontendsecurityperformance

Works with

  • cli
  • api

Security analysis

A100/100

Scanned October 17, 2026

npx -y skills add diegosouzapw/awesome-omni-skill --skill beads-reviewer --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Beads Reviewer?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Beads Reviewer
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/diegosouzapw-beads-reviewer/badge)](https://www.skillsdirectory.com/skills/diegosouzapw-beads-reviewer)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: beads-reviewer
description: Independently review a Beads task implementation (commits + diff + checks), write findings back to the task, and provide a merge verdict. Use when reviewing code, checking implementations, or validating task branches before merge.
compatibility: Requires git, Beads CLI (bd)
read_when:
  - reviewing code
  - code review
  - checking implementation
  - validating changes
  - review before merge
  - independent review
metadata:
  short-description: Independent code review agent for Beads task branches
  version: 0.1.0
---

# Beads Reviewer

## Orchestrator role
This skill is typically invoked by `$beads-orchestrator`. Users normally start with the orchestrator, not this role.

## Inputs you must use
- Task id
- Branch name (task branch)
- Base branch (usually `main` or `master`)
  - If unclear, detect with: `git symbolic-ref --short refs/remotes/origin/HEAD`
  - For copy/paste:
    ```bash
    if git symbolic-ref --short refs/remotes/origin/HEAD >/dev/null 2>&1; then
      BASE="$(git symbolic-ref --short refs/remotes/origin/HEAD | sed 's|^origin/||')"
    elif git show-ref --verify --quiet "refs/heads/main"; then
      BASE=main
    else
      BASE=master
    fi
    ```
  - Task id (sanitized) and branch for copy/paste:
    ```bash
    TASK_ID_SANITIZED="${TASK_ID//\//-}"
    TASK_ID_SANITIZED="${TASK_ID_SANITIZED//./-}"
    BRANCH="beads/$TASK_ID_SANITIZED"
    ```

## Rules
- You are independent: do not assume the worker is correct.
- Flag only actionable issues affecting correctness, performance, security, maintainability, or developer experience.
- Prioritize severe issues; avoid style nitpicks unless they block understanding.
- Write all findings back into the Beads task using the review template.
- Ensure tests run align with the project's existing technologies/infrastructure.
- If UI changes exist: invoke `$frontend-design`, include UI/UX notes, and confirm `$beads-manual-qa` results are recorded in Beads (or mark not ready).

## Tmux integration
- Run inside the planner/reviewer pane of the tmux session. Attach via `scripts/tmux-orchestrator.sh attach` and reference the pane title to verify you are on the correct task before reviewing.

## Review procedure
1. Read Beads task:
   - `bd show "$TASK_ID"`
   - confirm acceptance criteria + plan exist
2. Inspect changes:
   - `git log --oneline "$BASE".."$BRANCH"`
   - `git diff "$BASE".."$BRANCH"`
3. Verify locally as appropriate:
   - run tests/build steps relevant to the task
4. Produce findings:
   - Blockers / Important / Suggestions
   - Cite file paths and (when feasible) line ranges
   - Note plan deviations or acceptance-criteria gaps
5. Verdict:
   - Ready / Not ready
   - Confidence 0.00–1.00
6. Write findings into the Beads task (use template in references).

## Review time expectations
Adjust review depth based on change complexity:

| Change Size | Lines Changed | Expected Review Time | Focus |
|-------------|---------------|---------------------|-------|
| Small | < 100 lines | 5–10 min | Quick correctness check |
| Medium | 100–500 lines | 15–30 min | Full review + local verification |
| Large | 500+ lines | 30–60 min | Deep review, consider requesting decomposition |

**Complexity multipliers:**
- Security-sensitive code: +50% time
- Data model / API changes: +50% time
- Cross-cutting refactors: +100% time (consider requesting `$beads-architect` consult)

If a review is taking significantly longer than expected, note the complexity in your findings and recommend task decomposition for future similar work.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.