// TRUST · v0.5
tribunal
Adversarial multi-model code review with reputation-weighted findings settled on-chain.
The unit of trust is not consensus — it is surviving adversarial scrutiny by identified agents whose history is on the public record.
Three layers
1. Process backbone
State machine + spec-driven gates + role separation + git discipline. The mstar-harness lineage.
2. Correctness toolkit
Adversarial multi-model review on top of cooperative-parallel lens review, plus a verification pyramid. The colosseum lineage.
3. On-chain incentive layer
Soulbound reputation token on Burnt XION (CosmWasm) that tracks per-agent finding outcomes over time. The novel piece neither lineage had.
Composes with clawpatch
As of v0.3.4+, tribunal's lens-parallel review stage can run through clawpatch as a subprocess. The trust/discovery split:
- →clawpatch owns discovery — heuristic + agent-based feature mapping, per-feature LLM review, fix-and-revalidate loops.
- →tribunal owns trust — agent identity, ed25519-signed findings, adversarial multi-model review, PM/QA-settled outcomes, on-chain reputation.
"LLMs are fast, broad, and characteristically unreliable. Code review, tests, and audit are human-bottlenecked and scale linearly with reviewer attention. LLM output scales 10–100× faster. That mismatch is the trust gap."