Skip to content
roborev
GitHub stars

Agent Skills

Install slash commands that let AI agents fix review findings directly:

Terminal window
roborev skills install

Available Skills

SkillDescription
/roborev:address <job_id>Fetch a review and fix its findings
/roborev:respond <job_id> [message]Add a response to document changes

Usage

When you receive a review notification like “Review #1019: Fail”, ask your agent:

/roborev:address 1019

The agent:

  1. Fetches the review
  2. Reads relevant files
  3. Fixes issues by priority (high severity first)
  4. Runs tests
  5. Offers to commit

Agent-Specific Syntax

AgentSyntax
Claude Code/roborev:address, /roborev:respond
Codex$roborev:address, $roborev:respond

Updating Skills

Skills are updated automatically when you run:

Terminal window
roborev update

How It Works

Skills are installed as agent-specific configuration:

  • Claude Code: Custom slash commands in ~/.claude/
  • Codex: Custom agent skills directory

The skill scripts call roborev show --job <id> --json to fetch review data, then parse and present findings to the agent in a structured format.

Interactive vs Automated

Skills provide an interactive workflow - you ask your agent to address specific findings and review the changes before committing.

For automated fixing, use roborev refine which runs in a loop until all reviews pass.

See Also