Five layers, one defended agent.
AgentGuard is a Rust protocol behind DID/UCAN auth with per-tenant row-level isolation. Every layer below has a real, tested core; the evaluation, monitoring, cost and trust APIs are wireable today, and the on-chain layers (auto-pause + bounty escrow) go live once contracts are deployed + audited.
Pre-deployment Evaluation
Know an agent's risk before it ever touches a wallet.
A multi-LLM consensus engine generates novel adversarial scenarios tailored to the agent's category, scores how exploitable it is, and signs the report to IPFS.
What's under the hood
- ✓Dynamic scenario generation — not a static checklist
- ✓Multi-model consensus with confidence intervals
- ✓Category-aware (leverage trader vs NFT minter vs yield farmer …)
- ✓Cryptographically signed, verifiable report
Real-time Monitoring & Auto-pause
A firewall that halts a compromised agent mid-attack.
Three compute tiers — statistical baselines, a fitted robust anomaly model, and a budget-gated LLM explainer — flag a critical transaction and trigger enforcement: a signed off-chain pause directive, or an on-chain auto-pause once the agent is enrolled and the pause contract is deployed.
What's under the hood
- ✓Tiered detection: statistics → statistical ML → LLM
- ✓Robust anomaly model (dispersion-aware, drain-probe detection)
- ✓Semantic zero-day matching against a global threat corpus
- ✓On-chain auto-pause (or a signed off-chain directive) with a confidence score
LLM Cost Optimization
Cut an agent's inference bill; our fee is a share of the savings.
A strictly tenant-partitioned semantic cache reuses a prior response instead of re-calling the LLM. The fee model only charges on reported savings (per-agent savings metering + settlement are on the roadmap — computed today, not yet billed).
What's under the hood
- ✓Semantic LLM cache (pgvector KNN)
- ✓Strict per-agent + per-namespace isolation
- ✓15% fee model on reported savings — aligned incentives (metering + settlement pending)
- ✓Hard per-tenant compute-budget cap on every LLM call
Dynamic Trust Oracle
Public, evidence-aware reputation for every agent.
A live trust score with an honest confidence interval. A brand-new agent reads 50 ± 50 — not a fabricated certainty dressed up as fact.
What's under the hood
- ✓Evidence-aware margin — wide when data is thin
- ✓Security & reliability derived from real evaluation + monitoring data
- ✓Public and unauthenticated (aggregates only, never row data)
- ✓No spoofable inputs feeding the score
Vector-driven Bug Bounties
A paid zero-day hardens every agent, instantly.
Researchers submit exploits to an on-chain escrow. On payout the exploit is embedded into a global threat corpus, so the semantic matcher catches it across every agent and chain.
What's under the hood
- ✓On-chain bounty escrow marketplace
- ✓Paid exploits become global threat signatures
- ✓Chain-agnostic reuse — a Base zero-day protects Ethereum agents
- ✓Proof-of-Save engine turns each catch into a case study
Ready to guard an agent?
Start with a pre-deployment evaluation.