Overview — frameworks we surveyed
We surveyed every Claude Code agent framework we could find before deciding to write our own. Here’s the matrix.
| Framework | Author / Org | Stars (Apr 2026) | Core idea | Closest pattern we adopted |
|---|---|---|---|---|
| gstack | Garry Tan (YC) | 71K | 23 role-based slash commands, “boil the lake” | Boil the lake principle in 5 role prompts |
| Superpowers | Jesse Vincent | 149K | 7-phase TDD iron law | TDD iron law in validator |
| GSD | Lex Christopherson | 51K | Per-phase orchestrators with state-to-disk | Fresh contexts + cache-stable prompts |
| Agentwise | Phil | — | 8 specialist agents in parallel + dashboard | Real-time dashboard pattern |
| Hermes | (referenced in gstack) | — | Autonomous orchestration with checkpoints | Named checkpoints |
| Multi-Agent Ralph Loop | alfredolopez80 | — | MemPalace 4-layer memory | 4-layer memory (L0-L3) |
| ComposioHQ Agent Orchestrator | Composio | — | Git worktrees per parallel agent | Worktrees per lane |
| OpenSwarm | unohee | — | Linear-driven Worker/Reviewer pair | (rejected — no external task source) |
| Conductor | $22M Series A | — | Two-mode parallelism | Competition mode |
| claudecode-orchestrator | darrenapfel (deprecated) | — | “Quality through truth” + service smoke-test | Evidence rule + smoke-test gate |
| MOLTRON | — | — | Self-evolving Skills.md | TRICK: convention |
Plus general-purpose: AutoGPT, LangChain, CrewAI, MetaGPT, SuperAGI, Haystack, Semantic Kernel — these aren’t coding-specific and we didn’t borrow patterns from them directly.
Two ecosystem clusters
Section titled “Two ecosystem clusters”Reading the surveys, the Claude Code ecosystem clusters around two patterns:
- Skill packs. gstack, Superpowers, GSD — slash commands inside a Claude Code chat. Human is the orchestrator. Claude runs one task at a time. Useful for solo founders, but not autonomous.
- Multi-agent runtime. Agentwise, Hermes, Conductor, ComposioHQ — they run agents in parallel. Closer to what we needed, but each ties orchestration to their opinion of how the company should work.
We landed in a third cluster: a tiny supervised loop that owns no opinion about the company, just the agent runtime.
What we built ourselves
Section titled “What we built ourselves”Things no other framework had, that we needed:
- Decisions timeline with parser-ghost rate. None of them expose orchestrator parse-ghost telemetry.
- Cost-per-feature attribution via filename tagging. Others track per-session.
- Plan-reviewer role as an autonomous gate. gstack has the spiritual equivalent (
/plan-ceo-review) but human-triggered. - Autonomous Product role for scope expansion. Closest is MOLTRON’s self-evolving capabilities, but that’s capability evolution not scope expansion.
- Proposal CRUD + bulk accept/reject UI.
- Per-role identity files with curator-promoted patterns + worker TRICK: convention.
See What we kept, dropped, and invented for the synthesis.