Agentguard
This is a safety and governance layer for AI agents that sits between your LLM and the tools it calls. You get 20 tools covering policy checks before actions execute, risk scoring for operations, audit logging for compliance trails, and rate limiting to prevent runaway behavior. Reach for this when you're deploying agents in production and need guardrails around what they can do, how often, and with what level of scrutiny. It's policy as code for AI systems, letting you define rules centrally and enforce them across all tool calls without modifying individual integrations.
- tools
- 17
- transport
- HTTP
- last updated
- May 7, 2026
Tools
17 toolsPublic tool metadata for what this MCP can expose to an agent.
policy_preflightPre-flight check before any tool call. Evaluates 7 policies, computes risk score, detects threats, auto-logs. Returns allowed/denied/require_approval/flagged.
Pre-flight check before any tool call. Evaluates 7 policies, computes risk score, detects threats, auto-logs. Returns allowed/denied/require_approval/flagged.
No parameter schema in public metadata yet.
tool_risk_score0-100 risk score for tool + input. Detects secrets (API keys, passwords), prompt injection, high-value amounts.
0-100 risk score for tool + input. Detects secrets (API keys, passwords), prompt injection, high-value amounts.
No parameter schema in public metadata yet.
approval_requiredCheck if tool needs human approval. Optionally registers a pending approval request with tracking URL.
Check if tool needs human approval. Optionally registers a pending approval request with tracking URL.
No parameter schema in public metadata yet.
audit_log_writeWrite tool execution to persistent, cryptographically-signed audit log (SQLite WAL). Call after execution to record outcome.
Write tool execution to persistent, cryptographically-signed audit log (SQLite WAL). Call after execution to record outcome.
No parameter schema in public metadata yet.
audit_log_queryQuery audit trail. Filter by agent, tool, decision, time range. Paginated. Returns signed entries for tamper verification.
Query audit trail. Filter by agent, tool, decision, time range. Paginated. Returns signed entries for tamper verification.
No parameter schema in public metadata yet.
decision_explainHuman-readable explanation of any allow/deny decision. Pass request_id for stored entry or tool_name + tool_args for fresh analysis.
Human-readable explanation of any allow/deny decision. Pass request_id for stored entry or tool_name + tool_args for fresh analysis.
No parameter schema in public metadata yet.
rate_limit_checkCheck agent rate limits: 200/min, 5000/hr, 50000/day. Returns per-window usage with percentage.
Check agent rate limits: 200/min, 5000/hr, 50000/day. Returns per-window usage with percentage.
No parameter schema in public metadata yet.
payment_policy_checkValidate payment against policy: amount limits, recipient denylist, supported currencies/networks, AML thresholds, MiCA flags.
Validate payment against policy: amount limits, recipient denylist, supported currencies/networks, AML thresholds, MiCA flags.
No parameter schema in public metadata yet.
spend_limit_checkCheck per-call/hour/day spend limits by trust level. Default: 10k/call, 50k/hr, 200k/day. Trusted: 100k/call, 500k/hr, 2M/day.
Check per-call/hour/day spend limits by trust level. Default: 10k/call, 50k/hr, 200k/day. Trusted: 100k/call, 500k/hr, 2M/day.
No parameter schema in public metadata yet.
secret_exposure_checkDeep scan for 19 secret patterns: OpenAI/GitHub/AWS/Slack keys, Bearer/Basic auth, ETH private keys, Bitcoin WIF, credit cards, SSNs, emails. Returns severity + remediation.
Deep scan for 19 secret patterns: OpenAI/GitHub/AWS/Slack keys, Bearer/Basic auth, ETH private keys, Bitcoin WIF, credit cards, SSNs, emails. Returns severity + remediation.
No parameter schema in public metadata yet.
payload_safety_check18-pattern safety scan: prompt injection, jailbreak/DAN, role hijacking, SQL, XSS, Python/JS/Shell injection, path traversal, null bytes, oversized payloads.
18-pattern safety scan: prompt injection, jailbreak/DAN, role hijacking, SQL, XSS, Python/JS/Shell injection, path traversal, null bytes, oversized payloads.
No parameter schema in public metadata yet.
replay_guard_checkDetect replay attacks via SHA256 fingerprint (agent+tool+args). Configurable window (default 5 min). Returns duplicate count + first/last seen.
Detect replay attacks via SHA256 fingerprint (agent+tool+args). Configurable window (default 5 min). Returns duplicate count + first/last seen.
No parameter schema in public metadata yet.
cross_tool_anomaly_checkDetect anomalous patterns: risky combos, high frequency, repeated denials (policy probing), broad reconnaissance, elevated avg risk score.
Detect anomalous patterns: risky combos, high frequency, repeated denials (policy probing), broad reconnaissance, elevated avg risk score.
No parameter schema in public metadata yet.
scope_checkRole-based scope control. Roles: admin, compliance_officer, trader, auditor, developer, readonly. Returns has_scope, missing scope, granting roles. Logs denials.
Role-based scope control. Roles: admin, compliance_officer, trader, auditor, developer, readonly. Returns has_scope, missing scope, granting roles. Logs denials.
No parameter schema in public metadata yet.
session_validateFull session lifecycle: create (TTL + call budget), validate (increment counter), invalidate, info. Sessions carry role, scopes, tenant, expiry.
Full session lifecycle: create (TTL + call budget), validate (increment counter), invalidate, info. Sessions carry role, scopes, tenant, expiry.
No parameter schema in public metadata yet.
tenant_policy_checkMulti-tenant governance. Built-in tenants: default, fintech_eu, defi_protocol, enterprise_read. Per-tenant blocklists, risk limits, spend caps.
Multi-tenant governance. Built-in tenants: default, fintech_eu, defi_protocol, enterprise_read. Per-tenant blocklists, risk limits, spend caps.
No parameter schema in public metadata yet.
threat_intel_checkEntity threat intelligence. Auto-detects ETH addresses, IPs, domains. Checks sanctions, disposable services, behavioral analysis from audit log.
Entity threat intelligence. Auto-detects ETH addresses, IPs, domains. Checks sanctions, disposable services, behavioral analysis from audit log.
No parameter schema in public metadata yet.
Sponsors
Agent CaffeineSponsoredClose your lid, come back to finished work
SnugSponsoredThe clipboard manager for the AI eraPromote your product hereReach 125k+ visitors monthly
WaxSponsoredBlock anything with supervision, without erasing your iPhonePromote your product hereReach 125k+ visitors monthly
knobsSponsoredSee your site on any phone, any day, in any languageReadme
🛡️ AgentGuard MCP
Security, Policy & Audit Layer for AI Agent Tool Execution — 20 tools | Part of FeedOracle & ToolOracle
AgentGuard is the governance and security layer for AI agent workflows. Before any tool executes, AgentGuard evaluates policies, scores risk, detects secrets and injection attempts, logs to a tamper-evident audit trail, and explains every decision. Built for regulated environments, autonomous payments, and enterprise AI agent deployments.
Quick Connect
# FeedOracle (compliance-focused)
npx -y mcp-remote https://feedoracle.io/guard/mcp/
# ToolOracle (agent-commerce focused)
npx -y mcp-remote https://tooloracle.io/guard/mcp/
Details
- Repository
- github.com/tooloracle/agentguard
- Transport
- HTTP
- Registry
- MCP registry, active