AppSec
Runtime control evidence
Start with where stop, approval, and proof break or hold.
CAISI
Independent research and operating notes on AI Software Delivery Control.
CAISI Research
Artifact-backed findings for AI agent governance
Use this page when you need the measured claim, the artifact, and the scope limit before moving to interpretation.
Start by role
CISO / Security leadership
Approval, ownership, and evidence posture
Start with visibility, approval opacity, and governance proof.
Engineering / Platform
Operating model and delivery standards
Start with the repo, CI/CD, workflow, and proof standards that make AI-assisted delivery reliable.
GRC / Audit
Proof packets and evidence quality
Start with the evidence fields that can be reconstructed outside the product UI during audit, incident review, or customer review.
Published research
Published report
OpenClaw 2026
Governed vs ungoverned AI agent behavior in a controlled 24-hour run, with stop behavior, approval mediation, and evidence quality measured side by side.
Published report
AI Tool and Agent Sprawl 2026
A locked public GitHub cohort report focused on AI tool visibility, approval opacity, evidence posture, and governance readiness.
Archive and interpretation
Build archive
AI Tool Sprawl Q1 2026
The earlier build page for the flagship sprawl study remains available as background on the study's build phase and publication posture.
Interpretation layer
CAISI field notes
Field notes turn these reports into operating notes, role-specific lessons, and implementation guidance without changing the evidence base.