Control
Attack your own rules before anyone else does
Fire look-alike actions, missing fields and odd numbers at your policy. Get a grade, the exact gaps, and saved tests you can run on every change.
Gaps
- Amount missing, treated as £0Auto-approved
- £500.01 not handled by any ruleFalls to default
- Negative amount acceptedFalls to default
Saved as a test
- name: missing amount input: { action: refund.issue } expect: require 2 reviewersWhat you get
Policy Gauntlet, in practice
Finds the gaps you did not think of
Text where a number should be, a missing field, an action name with a space or a capital letter.
A grade and a fix for each gap
Every finding says what got through and what to change.
Policy tests as code
Save the cases as a file, then re-run them on a draft to see which answers change.
How it works
Step by step
- 1
Pick a policy
Use the live one or a draft.
- 2
Run the attack
Dozens of probes, built from your own rules.
- 3
Read the grade
Findings are ranked by how much harm they could do.
- 4
Save the tests
Re-run them whenever a rule changes.
Questions
About policy gauntlet
Does it touch my live policy?
No. It only reads a policy and reports. Nothing is saved unless you export the cases.
What does the mutation score mean?
We break your rules on purpose, one at a time, and count how many breakages your history would notice.
Next
Works well with
Know what your agents did.
Start free. Guard your first agent in a few minutes, and keep a record anyone can check.
- Free plan, no card
- Signed, chained decisions
- Checked offline, without us
Signed decision
- Decision
- Approved, 2 of 2
- Action
- refund.issue, £540
- Rule
- Two reviewers must agree
- Sealed
- 14:09, key 7f2c
- Hash
- 5be2…7c44
Checked on your own machine
- Signature matches the public key
- Decision is in the log
- Every link before it adds up