| Artifact | Category | Primary interface | Autonomy | Setup | Repo awareness | Best fit |
|---|---|---|---|---|---|---|
| AgentSys | Agent Skills | cli | multi-agent | high | deep | teams building repeatable agent pipelines |
| Claude Code Agents | Agent Skills | claude-plugin | agentic | medium | deep | solo founders who want prebuilt agent roles |
| Superpowers | Agent Skills | skills | agentic | medium | deep | teams wanting an explicit methodology |
| AB Method | Workflows & Knowledge Guides | slash-commands | agentic | medium | deep | incremental delivery |
| CCPM | Workflows & Knowledge Guides | skills | multi-agent | high | deep | spec-traceable team workflows |
| claude-code-tools | Tooling | cli | assistant | medium | medium | session continuity |
| Claude HUD | Status Lines | statusline | assistant | low | basic | tracking context health and agent activity |
| TDD Guard | Hooks | hooks | assistant | medium | medium | enforcing test-first discipline |
| Claude Squad | Tooling | tui | multi-agent | medium | deep | parallel task execution across isolated workspaces |
| Claude Code Chat | Tooling | vscode | assistant | low | deep | GUI-first coding with Claude Code |
Patterns visible in the matrix
- Methodology projects like Superpowers and AB Method optimize for disciplined planning and testing.
- Orchestration tools like AgentSys, CCPM, and Claude Squad favor worktree/session isolation and explicit coordination.
- Observability and enforcement tools like Claude HUD and TDD Guard are narrower, but they solve recurring operational pain.
- GUI/IDE tools like Claude Code Chat shift Claude Code into a more editor-centric workflow.
Prototype limitation
This page compares 10 enriched examples, not the entire 226-item corpus. The full seed table remains available in machine-readable form for later expansion.