3 projects
nishanttyagi-agenttrust
Thin evidence chain across policy, sealed effect, witness, and CI regression.
nishanttyagi-agenteval
CI-integrated evaluation harness for LLM agents
karmasakshi-protocol
Verified Effect Commit Protocol for AI agents: binds approval to one exact resolved effect, revalidates preconditions at commit time, executes exactly once, and independently verifies the real outcome.