Skip to content

Repository files navigation

AgentProof

AgentProof recommends which AI agent to hire using skill-specific trust evidence instead of a single global rating.

Start locally

Requirements: Node.js 20.19+ or 22.12+ and npm.

./start.sh

Open http://127.0.0.1:5173. The launcher installs locked dependencies when needed, runs the test/typecheck preflight, and starts the interactive demo. Use RUN_CHECKS=0 ./start.sh for a faster restart or PORT=5174 ./start.sh when the default port is busy.

The browser demo is local-first and fallback-safe. Optional Live tools checks run server-side through local /api/* routes and never expose HydraDB or RocketRide credentials to the frontend.

See WEBAPP_GUIDE.md for the detailed interaction flow, demo/live-service boundary, and troubleshooting steps.

Useful checks:

npm run typecheck
npm run build

Work independently

Start both feature branches from the boilerplate commit.

  • Mohit/Codex: use codex/hydradb-trust-layer; follow AGENTS.md; own hydradb/, src/server/hydradb/, src/server/trust/, and src/lib/trust/.
  • Teammate/Claude Code: use feature/rocketride-pipelines; follow CLAUDE.md; own rocketride/, src/server/rocketride/, and src/lib/rocketride/.

The files in contracts/ are shared. Coordinate before changing them. Each side can develop against rocketride/samples/trust-event.sample.json without waiting for the other service.

Read AgentProof_Project_Brief.md for canonical product context and AgentProof_Two_Person_Implementation_Plan.md for the full phased plan.

Repository map

contracts/        Shared JSON Schema integration boundary
data/seed/        HydraDB-side demo data
hydradb/          HydraDB setup, seed, and query artifacts
rocketride/       RocketRide pipelines and sample I/O
src/app/          Shared placeholder UI shell
src/server/       Server modules split by owner
src/lib/          Client/domain modules split by owner
src/types/        Shared and vendor-specific TypeScript types

About

AgentProof is the trust layer for AI agents that hire other AI agents. It decomposes a job into required skills, checks which agents have proven trust for those skills, and recommends whether to hire, warn, or avoid a candidate agent.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages