Action PoliciesHuman ApprovalsFull TracesKill SwitchCost AttributionAction PoliciesHuman ApprovalsFull TracesKill SwitchCost AttributionAction PoliciesHuman ApprovalsFull TracesKill SwitchCost AttributionAction PoliciesHuman ApprovalsFull TracesKill SwitchCost Attribution

Every agent action sentinel® approved.

A policy layer for teams shipping autonomous agents. Allow, deny, or hold every high-risk tool call for a human — with a full audit trail of what your agents did.

0%
Of agent actions audited
0
Decision verdicts: allow · deny · hold
0
Guardrails in the pipeline
0+
Providers with auto-fallback
0d
Default trace retention
How it works

Ask before acting.

Sentinel sits between your agent and the things it can do. Three steps stand between an agent's intent and a real side effect.

01

The agent asks first

Before running a high-risk tool, your agent calls the decision point — one HTTP request. Works with LangChain, CrewAI, MCP, or plain code.

02

Policy decides

Declarative per-tool rules — a glob plus a condition on the arguments — evaluate by priority. First match wins; unmatched calls fall through to your default.

03

You get a verdict

Allow, deny, or hold for a human — every decision recorded in an immutable audit log with who, what, and why.

The contract

Three verdicts, one endpoint.

ALLOW

No policy blocked it. The agent proceeds and the action is logged.

DENY

A policy — or the kill switch — forbids it. The agent is told it may not proceed.

HOLD

It crosses a line a human should see. The action waits in the approval queue.

Under the hood

On top of a full gateway.

The decision point rides on everything you already need to run agents in production.

Full traces

Every run is a span tree — prompt, guardrails, model, cost — PII-redacted at rest.

Guardrails

PII redaction and prompt-injection blocking before the model, leak-blocking after.

Provider fallback

Anthropic → OpenAI → Gemini. Kill the primary and the run still completes.

Cost attribution

Spend per tenant and per agent, with a monthly cap: block, warn, or degrade.

Kill switch

Freeze an agent and every action it attempts is denied instantly. No redeploy.

Eval CI gate

Score faithfulness and guardrail-pass-rate; block prompt regressions before they ship.

Govern your agents today.

Open source and self-hostable. One endpoint stands between your agent and a mistake it can't take back.

Start governing
POST /v1/agents/{id}/actions/check
{ "tool": "refund", "arguments": { "amount": 5000 } }

→ { "decision": "pending", "reason": "refunds over $100 need a human" }

Sits in front of the stacks you already build on

LangChainCrewAIAnthropicOpenAIGeminiLlamaIndexAutoGenMCPVercel AIPostgresLangChainCrewAIAnthropicOpenAIGeminiLlamaIndexAutoGenMCPVercel AIPostgresLangChainCrewAIAnthropicOpenAIGeminiLlamaIndexAutoGenMCPVercel AIPostgresLangChainCrewAIAnthropicOpenAIGeminiLlamaIndexAutoGenMCPVercel AIPostgres