RIGHT FIT
Take it before your agent acts on real data.
- Your agent can send, write, delete or pay
- It reads untrusted content: emails, documents, web pages
- Nobody has listed what it can actually do
Map your AI agent’s dangerous permissions in 90 minutes.
The Agent Permission Map is a $79 guided product for teams whose AI agent can take actions: write data, send messages, call APIs or move money. In four steps you list every tool, rank its impact, test misuse safely and decide a control for each action.
No code access neededAbout 90 minutesClear-outcome guarantee
Every toolListed with its credentials
ImpactRead to irreversible
Misuse testsInjection and loops, on staging
4 decisionsAllow, limit, approve, remove
The product
What is the agent permission map?
It’s a guided review of what your AI agent is allowed to do, before an injected instruction or a loop does it for you. You inventory its tools and data, rank each by impact, test misuse on staging, and choose a control for every action.
An honest boundary
Who needs an agent permission map?
Teams shipping an AI agent or assistant that calls tools: database writes, emails or SMS, payments, file access, browsing, code execution or MCP servers. You need access to the agent’s code and configuration and a staging environment.
NOT THIS PRODUCT
Choose another route for a different problem.
- A chatbot with no tools or actions
- You want it built or hardened for you: Agentic AI
- The wider app needs launch checks: the Launch Readiness Sprint
How it works
What are the four steps?
Four steps in about 90 minutes. Inventory every tool and data source, classify each by impact and autonomy, test prompt injection and tool misuse on staging, then build the map with a decision and control for every action.
Inventory tools and data
Every action, data source, trigger and credential the agent uses.
Classify impact
Read, reversible write, irreversible, external message, money or data leaving.
Test misuse safely
Prompt injection, tool misuse, loops and cross-user access, on staging only.
Decide the controls
Allow, allow with limits, require approval, or remove, per action.
Exact output
What do you get at the end?
An Agent Permission Map: every tool and data source with its credential, impact level, autonomy and misuse test result, and a decision (allow, allow with limits, require approval, or remove) plus the control to put in place for each.
Agent Permission Map
Every action with impact, autonomy and a decision.
Misuse test cases
Injection, tool misuse and loop tests you can rerun.
Control list
Scopes, approval gates, limits, allow-lists and logging to add.
Worked example
A completed map for a fictional agent.
After the product
What comes after the map?
The map tells you what to control. The Production-Safe Agent Sprint ($149) walks you through putting the controls in place, adding tracing and evals, and running failure drills. For an engineer review or implementation, start from your map with our Agentic AI team.
Low risk
What if the map isn’t complete?
You get your $79 back. Complete all four steps. If you don’t end with a map that has a decision for every action your agent can take, send your workbook within 30 days of purchase and we’ll refund you in full.
Clear-outcome guarantee
Complete the four steps and send the workbook within 30 days. If you don’t have a map with a decision for every action, we refund the full $79.
Stuck on a step?
Reply to any email from us with the step number. A Zenveus engineer answers within 1–2 hours during business hours. Steps only, not a code review.
Never wasted
If you later want Zenveus to do the work, the conversation starts from your output instead of from zero.
Straight answers
Frequently asked questions about the agent permission map
Do I test against real customers?
No. Misuse tests run on staging with test data and test credentials only.
Which agent frameworks does it cover?
Any agent that calls tools: custom code, OpenAI or Anthropic tool use, LangChain-style frameworks, MCP servers and builder-made agents.
Is this a security audit?
No. It’s a structured self-review of agent permissions. Specialist security work is separate.
Is there a refund guarantee?
Yes. Complete all four steps; if your map doesn’t have a decision for every action, send the workbook within 30 days for a full $79 refund.
The next decision
Know what your agent can do before it does it.
Four steps, every tool, one decision each.
Covered by the clear-outcome guarantee: complete the four steps, and if your map isn’t complete, get your $79 back within 30 days.