gan-planner
25Agent
GAN Harness — Planner agent. Expands a one-line prompt into a full product specification with features, sprints, evaluation criteria, and design direction.
Agent
GAN Harness — Planner agent. Expands a one-line prompt into a full product specification with features, sprints, evaluation criteria, and design direction.
Agent
Go build, vet, and compilation error resolution specialist. Fixes build errors, go vet issues, and linter warnings with minimal changes. Use when Go builds fail.
Agent
Expert Go code reviewer specializing in idiomatic Go, concurrency patterns, error handling, and performance. Use for all Go code changes. MUST BE USED for Go projects.
Agent
Analyze and improve the local agent harness configuration for reliability, cost, and throughput.
Agent
Plans evidence-backed knowledge structure strictly from a Domain Profile, Knowledge Blueprint, eligible Claim revisions, and declared templates.
Agent
Explains an approval-bound deterministic publication plan, including routes, change types, target conflicts, Context Pack registration, and rollback scope.
Agent
Independently reviews evidence-backed drafts and Agent Context Packs for unsupported statements, Verdict escalation, mapping errors, and retrieval gaps.
Agent
Operate autonomous agent loops, monitor progress, and intervene safely when loops stall.
Agent
Go/Python backend performance specialist. Use for profiling slow handlers, database bottlenecks, memory growth, goroutine/task leaks, inefficient algorithms, and production latency regressions.
Agent
Go/Python backend planning specialist. Use for features, refactors, migrations, bug fixes, and verification plans before implementation.
Agent
Expert Python code reviewer specializing in PEP 8 compliance, Pythonic idioms, type hints, security, and performance. Use for all Python code changes. MUST BE USED for Python projects.
Agent
PyTorch runtime, CUDA, and training error resolution specialist. Fixes tensor shape mismatches, device errors, gradient issues, DataLoader problems, and mixed precision failures with minimal changes. Use when PyTorch training or inference crashes.
Agent
Go/Python dead code cleanup and consolidation specialist. Use for removing unused code, duplicate helpers, stale imports, and low-risk refactors after tests are green.
Agent
Go/Python backend security reviewer. Use after changes involving user input, auth, APIs, database queries, files, secrets, payments, webhooks, or external services.
Agent
Reviews staged evidence-backed knowledge changes for semantic content not covered by the approved Synthesis Bundle and change IDs.
Agent
Review code for silent failures, swallowed errors, bad fallbacks, and missing error propagation.
Agent
Test-Driven Development specialist enforcing write-tests-first methodology. Use PROACTIVELY when writing new features, fixing bugs, or refactoring code. Ensures 80%+ test coverage.
Command
Agent system health check — audit Claude Code config, agent harness, context budget, autonomous loops, and workspace surfaces. Produces a prioritized remediation plan.
Command
Fix build, compile, lint, and type errors across Go, Python, PyTorch, and other stacks. Detects language automatically and delegates to the appropriate resolver agent.
Command
Build a full RAG knowledge base for the current repository. Scans all source code and generates .rag/ with layered docs (L0→L3), API contracts, ADR index, and GraphRAG knowledge graph. For small repos, runs fully automatically. For large/monorepo repos, use --large to enable multi-stage, checkpoint-resumable…
Command
Create, verify, or list workflow checkpoints after running verification checks.
Command
Database work — schema design, migration authoring, query optimization, and data-layer review across PostgreSQL, MySQL, and Redis. Delegates to database-reviewer agent.
Command
Architecture and API design — evaluate trade-offs with a four-voice council, apply hexagonal/backend patterns, design REST APIs, and record decisions as ADRs.
Command
Performance and error diagnosis — profile slow paths, detect silent failures, resolve build/runtime errors, and produce a root-cause report with fixes.