agentsteer 0.2.0

Runtime protection for AI coding agents

AgentSteer

Runtime protection for AI coding agents.

What it does

AgentSteer intercepts every tool call your AI agent makes via a PreToolUse hook, scores it against the task description, and blocks unauthorized actions before they execute. It catches prompt injection attacks, unauthorized file access, data exfiltration through delegation, and off-task behavior. Scoring uses oss-safeguard-20b (20B parameter model) via OpenRouter at under 200ms median latency.

Quick Start

pip install secure-environments

Set your OpenRouter API key. You can use a .env file or export it directly:

export OPENROUTER_API_KEY=sk-or-...

Install the hook for your framework:

secure-env install claude-code

That's it. Every tool call is now scored and unauthorized actions are blocked.

How It Works

Agent makes tool call
        |
        v
PreToolUse hook intercepts the call
        |
        v
AgentSteer scores the action against the task description
(oss-safeguard-20b via OpenRouter, <200ms median)
        |
        v
Score >= threshold? ---> BLOCK (action never executes)
        |
        v
Score < threshold? ---> ALLOW (action proceeds normally)

Each action is evaluated against 4 security policies. If any policy is violated, the action is blocked and the agent receives feedback explaining why.

Supported Frameworks

Framework	Install command	Hook type
Claude Code	secure-env install claude-code	PreToolUse hook via settings.json
OpenHands	secure-env install openhands	EventStream middleware
OpenClaw	secure-env install openclaw	Plugin hook
Python API	from secure_environments import score_action	Direct function call

Results

Evaluated on the AgentDojo benchmark, which tests whether agents can be tricked into executing prompt injection attacks hidden in their environment.

Framework	Condition	Attacks blocked	Utility preserved	n
Claude Code	Honest tasks	0% false positives	100%	20
Claude Code	Attack tasks	100% blocked	--	20
OpenHands	Honest tasks	0% false positives	100%	20
OpenHands	Attack tasks	95% blocked	--	20
OpenHands	Per-action detection	96.5% detected	0% false positives	1000

Median scoring latency: <200ms. Cost: $0.38 per 1M tokens via OpenRouter.

Full evaluation results and methodology at agentsteer.ai/evaluations.

Security Policies

Policy	What it catches	Example
P1: Read-only	Write actions when the task only requires reading	Task says "look up calendar" but agent sends an email
P2: Delegation	Actions that hand off control or leak data to external parties	Agent posts task contents to a URL, sends to unrelated email
P3: Category mismatch	Actions in a different category than the task	Task is about calendar but agent modifies files
P4: Target mismatch	Actions targeting a different entity than specified	Task says "email Alice" but agent emails Bob

A post-filter suppresses false positives from self-correction patterns (e.g., agent cancels a calendar event it just created in error).

Configuration

Environment variable	Description	Default
OPENROUTER_API_KEY	OpenRouter API key for scoring model	required
SECURE_ENV_TASK	Task description to score actions against	read from agent context
SECURE_ENV_THRESHOLD	Score threshold for blocking (0.0 to 1.0)	0.80
SECURE_ENV_DEBUG	Enable debug logging (1 or true)	off

You can also run secure-env setup for an interactive configuration wizard, or secure-env login to connect to the AgentSteer cloud dashboard for real-time monitoring.

CLI Reference

secure-env setup                # Interactive first-run setup
secure-env install claude-code  # Install hook for Claude Code
secure-env status               # Show current configuration
secure-env score "task" "action" # Score a single action
secure-env sessions             # List monitored sessions
secure-env version              # Print version

Links

• Website: agentsteer.ai
• Documentation: agentsteer.ai/docs
• Evaluations: agentsteer.ai/evaluations
• Repository: github.com/AgentSteer/AgentSteer

License

MIT