Latest, never best
The badge cannot cherry-pick a passing result. It always follows the most recent run and turns grey after 30 days.
- Latest run only
- Stale after 30 days
- 20-point score trend
- Publishing is reversible
Verifiable public proof
Turn the latest AgentLeak run into a public trust page and README badge. Visitors see the score, verdict, measurement date, detection tiers and trend—never the findings or sensitive values behind them.
The badge cannot cherry-pick a passing result. It always follows the most recent run and turns grey after 30 days.
A pattern-only pass is narrower than a full pipeline pass. The page names the tiers that ran, and a degraded run can never display the passing colour.
Public proof confirms that a measurement happened and how strong it was. It never exposes the findings, trace contents or sensitive values.
The mental model
A static badge can outlive the code it describes. AgentLeak binds the public claim to a dated latest run, makes weak or degraded detection visible and removes the passing state when the evidence becomes stale.
A badge is a permanent marketing claim.
A badge is a link to a current, qualified measurement.
How it works
Create the result you want the public claim to follow. The latest run will remain the source of truth.
Publish explicitly for the project; recording a run never makes anything public by itself.
Use the generated Markdown in a README, documentation site or release page.
Every new run refreshes the claim. Remove the slug at any time to withdraw the page and badge.
Publish and link the latest result
curl -X POST https://www.agentleak.org/api/projects/<id>/publish \
-H 'content-type: application/json' \
-d '{"slug":"support-bot"}'
[]
(https://www.agentleak.org/a/support-bot)FAQ
Ready when you are
Create a local workspace, run a bundled scenario, then wire AgentLeak into CI or let your agent onboard itself.