Find out what your agent can reach.
A fixed-scope review of what your agent can access, spend, and expose — before someone else finds out.
What we look at
- Tool and permission surface: everything the agent can call, and what happens when it calls it wrong
- Prompt injection paths through documents, web pages, and tool output
- Data flow: what leaves your environment, where it lands, and who can read it
- Spend controls and runaway-loop protection
- Evaluation coverage: what you are measuring versus what actually goes wrong
What you get
- A written report with findings ranked by severity
- Reproduction steps for each finding
- A prioritised remediation plan you can hand to your own team
- A walkthrough session
Common questions
The tool and permission surface, prompt injection paths, data flow out of your environment, spend controls, and evaluation coverage. In short, what the agent can reach, what it can spend, and what it can expose.
A written report with findings ranked by severity, reproduction steps for each one, a prioritised remediation plan, and a walkthrough session.
The audit is a fixed-scope review and delivers a remediation plan your own team can act on. Implementation work is scoped separately.
It is the usual starting point when you are not certain what you have. The scope and price are fixed, and the output is something you can act on regardless of who does the work.
Tell us what you are shipping.
Send the shape of the problem — what the agent does, who uses it, and what breaks today. We will tell you what we would run and what it would take.