v1.4.0 · 1,400+ tests · Apache-2.0

The open-source
AI agent firewall.

Write a CEL rule. Strathon blocks the tool call before it executes, in-process, in under a millisecond. Three lines of code, ten frameworks, self-host in minutes.

Get Started GitHub
SOC 2 pathGDPREU AI Act evidenceOWASP ASI Top 10
agent.py · strathon
$ pip install strathon
Successfully installed strathon-1.4.0
from strathon.instrumentation.langgraph import instrument
handler = instrument(client)
agent.invoke(msg, config={"callbacks": [handler]})
✗ StrathonPolicyBlocked: tool 'drop_table' denied
policy="block-destructive" action=BLOCK <1ms
Agent
Tools / LLM
STRATHON
1,400+
tests passing
10
framework integrations
7
enforcement actions
v1.4.0
latest release

Built on open standards, no black boxes

OpenTelemetryCEL (Google)PostgreSQLOWASP ASIDockerApache-2.0
Three lines. Every framework.

Wrap your agent. Keep your code.

agent.py
from strathon import Client, instrument

client = Client(api_key="stra_...")  # endpoint defaults to localhost:4318
instrument(client, frameworks=["crewai"])
# tool calls now pass the policy engine before executing
See it block

Send an attack. Watch the verdict.

BLOCK
Tool call denied before execution
policyblock-prompt-injection
latency<1ms
verdictprompt_injection / tool_abuse
Capabilities

A control plane for everything your agents do.

CEL Policy Engine

Seven enforcement actions, 12 OWASP templates, and a shadow mode that tests policies against live traffic without blocking it.

Human Approval

Pause agent execution until an operator approves. Multi-party sign-off for high-stakes actions.

70+ Credential Patterns

Detect AWS keys, GitHub tokens, Stripe keys, private keys and more, redacted at ingest.

EU AI Act Evidence

Evidence export for Articles 9–15 & 19. NIST AI RMF risk scoring built in.

Behavioral Drift Detection

EWMA/CUSUM statistical analysis flags when an agent silently changes behavior.

Dashboard

Trace waterfall, policy editor, approval cards, audit log, and budget charts in one operator console.

The dashboard

Every agent action, on one screen.

app.getstrathon.com / traces
Traces
Policies
Approvals
Agents
Trace · checkout-agent312ms · 6 spans
agent.plan
llm.completion · gpt-4o
tool.search_docs
tool.read_file
tool.send_email
BLOCKED
agent.respond
Span detail
BLOCK
toolsend_email
policyblock-pii-exfil
latency<1ms
actionBLOCK

A live trace waterfall of tool calls and model invocations, with the exact policy verdict on every span, including what got blocked and why.

Works with the agent stack you already use

LangGraph
CrewAI
OpenAI Agents SDK
OpenAI
Anthropic
LangChain
AutoGen
Claude Agent SDK
Pydantic AI
Google ADK
LangGraph
CrewAI
OpenAI Agents SDK
OpenAI
Anthropic
LangChain
AutoGen
Claude Agent SDK
Pydantic AI
Google ADK
Sub-ms
in-process policy overhead
In your process
enforces at the tool-call boundary
Horizontal
stateless receivers scale out
Fail-closed
matched policies block when enforcement is incomplete

Throughput depends on your hardware and database, so Strathon ships a reproducible benchmark you run yourself rather than quoting a single number. See the scaling guide.

Threat model

Anchored on the OWASP Top 10 for Agentic Applications (2026).

ASI01
Agent Goal Hijack
CEL policies on prompt and input content; block or alert on hijack attempts
ASI02
Tool Misuse and Exploitation
Deny-by-default tool allow-lists, CEL argument rules, approvals for sensitive tools
ASI03
Identity and Privilege Abuse
Scoped API keys, RBAC, MFA, per-key rate limits
ASI04
Agentic Supply Chain Vulnerabilities
MCP gateway policy checks, egress domain allow-listing, credential scanning
ASI05
Unexpected Code Execution
Block and allow-list policies on shell, code, and SQL tools; approval before execution
ASI06
Memory and Context Poisoning
Drift detection surfaces poisoning effects; halt propagation; content redaction at ingest
ASI07
Insecure Inter-Agent Communication
MCP gateway evaluates inter-agent and tool calls; fails closed when evaluation cannot complete
ASI08
Cascading Failures
Cost and iteration budgets with auto-halt, circuit breakers, kill switches, halt propagation
ASI09
Human-Agent Trust Exploitation
Multi-party (N-of-M) approval, tamper-evident audit log, SARIF export
ASI10
Rogue Agents
Behavioral drift detection (Vigil), heartbeat monitoring, SDK integrity check, kill switches
Full OWASP mapping Agent firewall checklist
Self-host in minutes

Free forever, on your own infrastructure.

No license keys. No phone-home. Clone the repo, run Docker Compose, open the dashboard, and create your first policy. The full firewall (every scanner, every framework integration) is open source.

Read the quickstart
terminal
git clone https://github.com/strathon/strathon.git
cd strathon
docker compose up -d

# open localhost:3000: register, create a policy, done
Shipping fast

Latest releases.

View all
v1.4.0Hardening release2026-07-31
  • Every enforcement surface now fails closed when it cannot confirm a call is allowed. An enforcing policy whose expression cannot be evaluated blocks the call rather than letting it through, and the full policy set is evaluated first, so one policy's error no longer drops another policy's match on the same span.
  • The MCP gateway and egress proxy fail closed: the gateway validates the upstream address and ignores any caller request to turn its enforcement off, and the egress proxy blocks until it has loaded the project's policy, rather than starting open.
  • Every SDK adapter fails closed when the receiver is unreachable, and policies are evaluated against the full tool arguments, so a padded argument cannot slip content past a rule.
  • Project and member operations are confined to the caller's organization, closing cases that could read or act across tenants. Creating an API key or deleting a project now requires re-authentication.

Start protecting your agents.

Ship a firewall in front of every tool call today. It's open source. You can read every line.