Bmad Reviewer Lite — Development agent for Claude Code
Service type for the bmad-build-auto pipeline — the cheapest review layer; only the Intent Alignment Auditor gets it (on --fast and --normal).
How to install Bmad Reviewer Lite
Installs to ~/.claude/agents/staspotapov-auto-bmad-bmad-reviewer-lite.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/StasPotapov/auto-bmad/HEAD/agents/bmad-reviewer-lite.md -o ~/.claude/agents/staspotapov-auto-bmad-bmad-reviewer-lite.md Restart Claude Code, or start a new session, for it to be picked up.
What Bmad Reviewer Lite does
name: bmad-reviewer-lite description: Service type for the bmad-build-auto pipeline — the cheapest review layer; only the Intent Alignment Auditor gets it (on `--fast` and `--normal`). Do NOT pick it for ordinary tasks and do not use it instead of /code-review. Model Haiku 5.5, effort high: it fits a closed, descriptive "intent against diff" comparison with no repository walk and no bug hunting. Do not give this type to layers that look for defects or walk the tests. It is started by an inst
Alternatives in Development
- Surface Discovery Agent — Runs bounded normal surface-discovery — subdomain enum, live hosts, archived/crawled URLs, nuclei, JS/JWT extr 97 ★
- Bmad Builder 89 ★
- Po Evidence — Evidence seat on the Atlas PO Council 79 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Bmad Reviewer Mid
Service type for the bmad-build-auto pipeline — a diff review layer at the --normal / --fast level. Do NOT pic
Bmad Compiler Lite
Service type for the bmad-build-auto pipeline — builds epic- -context.md from the epics file and the planning
Miner Low
Delegate worker pinned to LOW reasoning effort, the cheapest tier. Pick for mechanical retrieval, extraction,
PR Manager Lite
Trimmed variant of pr-manager for fast iterations. Reads reviewer comments on a Gemma-Code PR, applies in-scop
Change Tracker
Identifies Intent Layer nodes affected by code changes for targeted validation
Kg Engineer
Senior Python Engineer for Knowledge Graph integration with CognivAgent. Specializes in Claude Agent SDK MCP t
Related Skills
Test Review
Write-then-verify test review — write tests in your normal flow, then mechanical gates (red-green + mutation)
Occams Razor Auditor
Claude Skill built for one specific move: catching a parsimonious model at the moment it gets stretched past t
Intent Drift Skill
Detects intent drift in AI-assisted development by comparing original goals against current execution plans us