👨‍💻 Code Reviewer

Mode: code-reviewer · Category: Development

Pull request reviews, refactor suggestions, code style critique

Quick run

From the repo root, after pnpm -r build:

node apps/cli/dist/index.js -a code-reviewer -q "your question here"

Requirements

LLM provider

Recommended: anthropic, openai gpt-4-turbo, deepseek-coder

Set the corresponding API key in ~/.openvesper/.env:

  • ANTHROPIC_API_KEY (or other)
  • GITHUB_TOKEN (for PR operations)

Plugins used

This agent has access to all plugins (cross-plugin tool registry), but typically reaches for:

filesystem, code, github

Skills it can pull in

bug-triage, code-explanation, debug-prod-issue, pr-description, pr-review-checklist, refactor-plan

Example sessions

node apps/cli/dist/index.js -a code-reviewer -q "Review my recent changes in apps/cli/src/"
node apps/cli/dist/index.js -a code-reviewer -q "List open PRs in openvesper/openvesper older than 3 days"
node apps/cli/dist/index.js -a code-reviewer -q "Write a PR description for the changes I just staged"

Scheduled mode (optional)

This agent ships with a heartbeat checklist in.agents/code-reviewer/HEARTBEAT.md. The default schedule is0 9 * * * but the heartbeat is disabled by default.

To enable scheduled execution:

# Edit .agents/code-reviewer/HEARTBEAT.md
# Change "enabled: false" to "enabled: true"

# Then add it to your cron job state
node apps/cli/dist/index.js cron add hb-code-reviewer   --schedule "0 9 * * *"   --agent code-reviewer   --prompt "Run your heartbeat checklist for {{date}}"   --deliver-to "telegram:@me"

Files

The agent's source files live under .agents/code-reviewer/:

  • SOUL.md — persona and principles
  • IDENTITY.md — metadata (name, icon, model preferences)
  • USER.md — what the agent knows about you
  • TOOLS.md — tool policy
  • HEARTBEAT.md — scheduled checklist (opt-in)
  • MEMORY.md — long-term memory store

You can edit any of these files to customize the agent. Changes take effect on the next run — no rebuild needed.

Privacy

This agent runs entirely on your machine. Your prompts go only to your configured LLM provider. Nothing is sent to OpenVesper servers (we don't have any). SeeSecurity policy for details.

What's next?