8
agent frameworks supported on day one
OTel
native instrumentation, no parallel pipeline
<1ms
capture + sign overhead per call
100%
tool calls and sub-agent spawns signed
What the SDK does

Three responsibilities. One library. One config block.

The Sentinel SDK runs in-process inside your agent framework. It captures every tool call, sub-agent spawn, and internal prompt via OpenTelemetry, before execution. It signs every agent session with a verifiable identity. And it enforces Kill Switch isolation at the call boundary, before any tool call leaves the process. Integration is one decorator and a config block.

01 / Observe

Capture every tool call and sub-agent spawn, before execution.

The SDK hooks the agent framework's tool-call dispatch. Every call is captured as an OpenTelemetry span with full parameter context, then routed to Time Machine and Overseer in parallel with execution. If a policy decision is required (rare; most policy runs at the Gateway), the call is held until the verdict returns.

  • OpenTelemetry-native. Spans flow through your existing OTel collector. No parallel pipeline.
  • Pre-execution capture. Tool call observed before the call leaves the process, not after, when blocking is no longer possible.
  • Sub-agent lineage. When an orchestrator spawns a child, the child's spans link to the parent and to the original human session.
python · langchain integration
from bladerun import Sentinel

# one decorator on your agent factory
@Sentinel.govern(agent_id="payment-orch")
def build_agent():
  return AgentExecutor(
    agent=agent,
    tools=tools,
    verbose=True,
  )

# every tool call now flows through
# Sentinel before execution.
02 / Sign

Sign every agent session with a verifiable cryptographic identity.

When an agent session starts, the SDK requests a signed session credential from the local identity provider, your existing AP2, Okta, Ping, or Entra installation; X.509 fallback if needed. Every outbound call carries the signed identity. The Gateway rejects unsigned traffic at L1. The audit trail is signed end-to-end.

  • Verifier, not issuer. Sentinel verifies whatever identity your organization already uses. We do not introduce a parallel identity system.
  • Per-session credentials. Short-lived signed tokens with the originating human principal embedded.
  • Tool-permission certificate. The agent's allowed tool set is signed alongside the session credential, enforced at Gateway L2.
session signing flow
# 1. agent session starts
session = Sentinel.begin_session(
  principal="jane.doe@bank",
  agent_id="payment-orch",
  tools=["read_account", "summarize"],
)

# 2. signed credential issued
session.credential
# > ed25519:sess_8a91d2c1f4...

# 3. every outbound call carries it
# Gateway L1 verifies. Unsigned = reject.
03 / Enforce

Subscribe to the Kill Switch. Enforce isolation at the call boundary.

Every Sentinel SDK in the agent fleet maintains a Redis pub/sub subscription. When Overseer fires the Kill Switch on a session, the targeted SDK denies any further tool call within milliseconds, at the SDK boundary, before the call leaves the process. Median time from anomaly detection to total isolation: under 50ms.

  • In-process enforcement. Isolation does not depend on the Gateway being reachable. The decision happens inside your agent.
  • Redis pub/sub fan-out. Single isolation event reaches every Sentinel in the fleet within ms.
  • Manual + automatic. SOC dashboard can fire isolation manually; Overseer fires it automatically. Same mechanism.
isolation flow
# overseer detects anomaly · score 0.91
# > KillSwitch.fire(session_id="sess_8a91")

# every Sentinel SDK sees it within ms
# via redis pub/sub fan-out

# next tool call inside that session:
agent.invoke("transfer 50k to...")
# > SentinelIsolatedError(
# >   session_id="sess_8a91",
# >   isolated_at="2026-04-18T14:22:13Z",
# >   reason="overseer.anomaly_score_threshold")
Framework support

Eight frameworks supported on day one. More on a published cadence.

If your agent framework emits OpenTelemetry spans, Sentinel SDK integrates via configuration. For frameworks without native OTel support, we maintain instrumentation libraries that ship and update independently of the framework's release cycle.

LangChain GA

Python · TypeScript

LangGraph GA

Python · TypeScript

LlamaIndex GA

Python

AutoGen GA

Python

CrewAI GA

Python

Semantic Kernel GA

Python · .NET

OpenAI Agents SDK GA

Python · TypeScript

Anthropic Claude SDK GA

Python · TypeScript

Bedrock Agents BETA

Python · Java

Vertex Agent Builder BETA

Python · TypeScript

Spring AI Q3 2026

Java

OTel-only (custom) GA

Any language · OTel SDK

Where the SDK lives

Inside your agent process. One library. Three downstream connections.

The SDK is a library, not a service. It runs in your agent's process, in your container, on your infrastructure. The three downstream connections, Gateway (outbound calls), Time Machine (logging), Kill Switch (isolation), are configurable per environment and fail closed by default.

Your agent process

Sentinel SDK

In-process. Hooks the framework's tool-call dispatch. Captures, signs, enforces.

Outbound calls

Gateway

Every signed outbound call routes through the Gateway for inspection.

Log + enforce

Time Machine + Kill Switch

OTel spans land in Time Machine. Redis pub/sub feeds isolation events.

Works with

The Sentinel SDK is one of six platform components.

Common questions

What engineering teams ask first.

How invasive is the integration?

A single decorator on the agent factory and a config block in your environment. The SDK hooks the framework's existing tool-call dispatch. Your agent code does not change. For frameworks with strong OpenTelemetry support (LangGraph, LlamaIndex, OpenAI Agents SDK), the integration is configuration-only.

What's the per-call overhead?

Under 1ms for capture and signing. The SDK runs in-process and uses an in-memory ring buffer for span emission; OTel export happens asynchronously. Kill Switch checking is a hot-path Redis call against a local-region replica, sub-millisecond at p95.

Does the SDK introduce a new identity provider?

No. Sentinel verifies whatever identity your organization already uses, AP2, Okta, Ping, Entra, and falls back to X.509 only if needed. We do not introduce a parallel identity system. The session credential is signed by your existing identity infrastructure; Sentinel produces and propagates it.

What happens if Redis is unreachable when the Kill Switch fires?

The SDK falls back to a local cache of recent isolation events refreshed on a 1-second cadence. In a worst-case scenario, the maximum window during which an isolated agent could complete additional calls is approximately the cache refresh interval. The configuration of this cache is exposed; tighter intervals are supported for the highest-stakes agent classes.

Can the SDK run in air-gapped environments?

Yes. In air-gapped deployments, the Kill Switch backplane runs locally and the SDK subscribes to it the same way. Federation participation is the only feature that requires connectivity, and it is opt-in. The SDK and the rest of the platform are fully functional offline.

What's the deployment story for sub-agents and orchestrators?

When an orchestrator spawns a sub-agent, Sentinel propagates a child-session credential that links back to the parent. Tool-call permissions on the child are scoped down by the parent at spawn time. The child cannot exceed the parent's tool certificate. Time Machine reconstructs the full lineage, and Overseer scores the orchestrator and each child independently.

Integrate the SDK

One decorator. One config block. Every action signed.

Pilot the SDK in one agent. OTel spans flow to your existing observability stack alongside Time Machine. Visible results without any blocking enforcement until your team is ready.