Runs in your browser. Nothing you enter is uploaded or logged.
Agent permission mapper
See how an attacker could abuse your agent's tools, and how to lock it down.
One useful email every Monday
New free tools, AI and network threats that matter, and one quick challenge. No spam, unsubscribe in one click.
How to use it
- 1
List your agent tools and permissions.
- 2
See the abuse paths.
- 3
Apply the recommended limits.
Questions people ask
What is excessive agency?
An AI agent with more tools, permissions or autonomy than it needs, so a prompt injection can make it do real damage.
Related tools
- LLM app threat modelPick your architecture and get the threats and controls that apply.
- System prompt hardenerScore your system prompt against common attacks and get a hardened version.
- Red-team test generatorSafe, canary-based injection tests for your own chatbot or agent.
- Agent trace analyzerFind goal hijacking, exfiltration and dangerous commands in agent logs.
- Guardrail config generatorOne policy for Bedrock Guardrails, Azure Content Safety, NeMo and Llama Guard.
Link to this tool
Found it useful? Paste this into your blog, wiki or runbook.
<a href="https://toolbini.com/ai/agent-permission-mapper/">Agent permission mapper</a>Reviewed by the editorial team · Last reviewed 2026-11 · Report an error