#agent-safety

Local-first memory firewall and MCP server for AI agents. Audrey gives coding agents recall, memory capsules, preflight checks, reflexes, validation, tool-trace learning, contradiction handling, and SQLite/sqlite-vec storage.

Behavioral trust scoring and MCP gateway proxy for 14,820+ MCP servers. Query real attestation data — interaction history, success rates, latency, uptime — before your agent executes sensitive operations. Gateway proxy routes agent calls through Observatory for automatic trust checks and signed attestation receipts. Compliance tier generates MiCA Article 12 audit trails for EU regulatory requirements (enforcement July 1, 2026). Supports Stripe metered billing and x402 USDC micropayments. 8 tools for trust evaluation, server discovery, anomaly detection, and compliance reports.

Make your AI agents cheaper, faster, and more reliable. 5 products, 14 MCP tools, 1 API key. Task deduplication, cost prediction, output validation, behavioral governance, and shared failure intelligence. Anonymous demo mode — try all tools without a key. 1,000 free credits/month, no card required. https://plith.ai/.well-known/mcp/server-card.json

PR-native API governance for AI agents. Detects breaking API changes at merge time, scores blast radius, and blocks unsafe agent tool calls before production. Preflight check, risk scoring, Decision Spec v1.0.

Security scanner for AI Agent skills, plugins, and MCP servers. 39+ detection rules with AST taint tracking, A-F grading, and post-uninstall residue cleanup.

Stateless circuit breaker that detects recursive agent loops in MCP tool calls. Uses Bloom filter to identify repeated calls within 30 seconds. Pay $0.005 USDC via x402 on Base network to reset context and break the loop.

Deterministic MCP server that scores code risk BEFORE execution. Returns ALLOW or BLOCK with full reasoning. Same inputs → same output. Always.