Use when you want a code review on recent changes, a specific file, or a pull request — 'review this', 'check my code', 'what did I miss'. Works in any project without setup.
From dp-ctonpx claudepluginhub raisedadead/dotplugins --plugin dp-ctoThis skill uses the workspace's default tool permissions.
Designs and optimizes AI agent action spaces, tool definitions, observation formats, error recovery, and context for higher task completion rates.
Implements structured self-debugging workflow for AI agent failures: capture errors, diagnose patterns like loops or context overflow, apply contained recoveries, and generate introspection reports.
Compares coding agents like Claude Code and Aider on custom YAML-defined codebase tasks using git worktrees, measuring pass rate, cost, time, and consistency.
Standalone code review skill. Works in any project without beads or an active dp-cto epic. Bypasses stage enforcement via the adhoc-* prefix.
Determine what to review:
git diff --name-only HEAD~1 has changes → use the diffPresent a lens menu via AskUserQuestion:
If the user provided a lens in their args (e.g., /dp-cto:adhoc-review security), skip the menu.
Spawn reviewer agent(s) via Agent tool with subagent_type: "dp-cto:reviewer".
For Quick: 1 agent with the selected lens.
For Full: 2 agents in parallel (primary lens + adversarial lens).
For Sweep: 1 sweeper agent instead.
For Pushback: 1 agent with pushback lens.
Include in the prompt:
Parse agent receipt(s). Present findings grouped by severity:
If multiple agents ran, deduplicate by file:line (keep higher severity).
This skill operates without beads. No epic, no task tracking, no stage transitions. The adhoc-* prefix ensures the skill-gate hook passes it through at any stage.