#evidence

Open-source loop engineering for auditable business briefings: claims, evidence, gates, findings, repairs, and human review.

Local-first research control plane for gated source packs, claim truth, contradiction handling, and long-running AI synthesis

A governed world-model evidence layer for AI agents: simulate bounded scenarios, track assumptions, score prediction-vs-reality error, and produce human-reviewable execution evidence.

Capability Host Protocol — make agent and tool execution visible, replayable, and ready for governance.

Guarded CLI, TypeScript client, and MCP server for JustHandled x402 agent utilities

Local-first evidence console for AI-assisted software work

Local-first agent operating system for durable AI coding: task graphs, deterministic context packs, executable delegation, verification evidence, and approval gates.

Require exact PR evidence before agent-authored changes merge. Local evidence MCP + base-trusted gate.

Find how a robot fails. Repair it. Prove it. Failure discovery → repair → statistical qualification for contact-rich manipulation, end to end on one AMD Radeon GPU — with a tamper-evident evidence bundle a judge verifies on CPU.

GitHub Action + CLI for replayable CI/PR/release gate decisions - zero-config Self-Test turning checks, scanners, and AI-agent signals into deterministic, tamper-evident PASS/WARN/BLOCK records. Read-only, zero dependencies, Apache-2.0.

Auditable signal cards and evidence ledgers for financial research agents.