Public audit library

AI Delivery Risk Audit — three public repositories, one method

Seven checks for the structural failures that agent-built codebases share. Every judgment carries a file path or a count you can re-run.

Run the same seven checks on your own repository. Free, read-only, about 15 seconds, no signup.
Scan a repo →
What the method found on mature, well-funded repositories

The audits

RepositoryScoreRisk levelWeakest check
pydantic/pydantic-ai80.2MediumComplexity Control (1/10)
openai/openai-agents-python77.8MediumComplexity Control (1/10)
microsoft/autogen55.2HighComplexity Control (2/10)

The seven checks

#CheckThe question it answersWeight
C1Scope ContainmentWhen the agent edits a file it should not touch, what stops it?18%
C2False-Green DetectionCan your tests tell “it ran” apart from “it is correct”?18%
C3Drift ContainmentIs “done” written down where the agent can read it?14%
C4Memory IntegrityWhen two sessions disagree, who arbitrates?10%
C5Delivery GateWhat blocks 80% done from being reported as done?16%
C6Complexity ControlAre files so large the agent’s context truncates mid-edit?12%
C7Regression SurfaceAfter one change, can you enumerate what else moved?12%

Risk bands: 85+ low · 70-84 medium · 50-69 high · below 50 severe.

Run it on your own repository

Three ways in. The free scan scores a public repository in about fifteen seconds, no signup. The self-serve report is the full write-up you run yourself. The audit is 7 days, human-verified, with a recorded walkthrough of your code and a free re-scan after your fixes.

Free scan — $0 Full report — $297 7-day audit — $2,500

Not sure which one? What the seven checks measure, what each tier includes, and the questions buyers ask →