Agent governance and control-plane systems, open source where public, described where private.
110-agent finance platform where a deterministic policy engine runs tenant, capability, and approval checks before any LLM call. $0 default autonomous spend, database-enforced immutable audit log.
Live demo and public repo. Last verified 2026-08-07.
Proves the mechanism behind the Systems Decision Audit at TechTide AI.
Trust contracts as JSON Schema: risk tiers, tool permissions, deny-by-default egress, approval gates, compliance mappings. Prompt-injection threat model with adversarial fixtures, ~28 CI gates.
Live docs and public repo. Last verified 2026-08-07.
Proves the mechanism behind the Systems Decision Audit at TechTide AI.
Multi-agent debugger with zero LLM calls in the analysis path: causal graphs, weighted root-cause scoring, failure prediction. 206 zero-mock tests, zero runtime deps, no network egress.
Live site and public repo. Last verified 2026-08-07.
357-agent orchestrator governed by per-agent budget caps, a pattern-matched bash security gate, auto-downgrade daily budgets, a hard per-run spend ceiling, and correlation-ID telemetry.
Live site and public repo. Last verified 2026-08-07.
Time-travel debugging for agent traces: zero-trust import parsing with signature verification and size clamps, five trace formats auto-detected, and a divergence/latency/token diff engine.
Private build. Technical writeup published. Last verified 2026-08-07.
Deep-research agent with a real human-in-the-loop approval workflow, where the run suspends until a human approves, plus a deterministic citation audit with zero AI in the verification path.
Public repo. No hosted demo. Last verified 2026-08-07.
Governed agent runtime: kill switch with cost and error auto-triggers, circuit breakers, a four-tier execution sandbox, hard daily and per-task budget caps, and a dead-man's switch.
Private build. Technical writeup published. Last verified 2026-08-07.
Proves the mechanism behind the Systems Decision Audit at TechTide AI.
Company-scale agent platform on a three-plane architecture covering control, data, and evidence. State-machine status transitions, hard execution timeouts, adversarial eval harness with a 33-task golden suite.
Public repo. No hosted demo. Last verified 2026-08-07.
Proves the mechanism behind the Systems Decision Audit at TechTide AI.
If an AI workflow works in demos but has not earned trust in operations, send me the workflow and what is blocking the call.
I usually reply within 24 hours.
Repo
