Control evidence for AI coding agents
Agents write the code.
Your controls still apply.
What the agent changed. Who approved it. The evidence your auditor asks for.
- 09:41:02session.startedclaude-code · payments-api
- 09:41:09intent.bundle.attachedspec PAY-2141 · sha256:a41f…
- 09:41:33tool.startedEdit src/ledger/settle.ts
- 09:41:34approval.requestedconfirm · payments-write
- 09:42:10approval.decidedapproved · second reviewer · rationale
- 09:43:05hook.blockedBash curl → public host
- 09:43:41model.completed1 secret redacted before emit
- 09:46:01session.finishedPR evidence digest sha256:f9b0…
On the record from
Capture status01Industry trend
Adoption outpaces governance.
83%
of developers use AI tools or agents at work
Stack Overflow 2026 (opens in a new tab)Oct 2026 · n = 17,464
24%
say their organization provides approved AI tools and guidance for using them
Stack Overflow 2026 (opens in a new tab)Oct 2026 · n = 13,167
6.6%
trust AI output for important work decisions
Stack Overflow 2026 (opens in a new tab)Oct 2026 · n = 14,304
45%
of test cases where AI models introduced an OWASP Top 10 vulnerability
Veracode 2025 (opens in a new tab)Jul 2025 · 80 tasks, 100+ models · vendor research
2 Dec 2027EU AI Act rules for Annex III high-risk systems apply, including logging (Art. 12) and human oversight (Art. 14). European Commission, AI Act (opens in a new tab)
Sources: Stack Overflow Developer Survey 2026 (self-reported; data under ODbL 1.0) and the Veracode GenAI Code Security Report. AutoDevOps is not affiliated with either publisher.
02How it works
Questions your auditor will ask.
63 controls · 18 frameworksWho approved the agent’s change?
The reviewer, the decision, and the rationale.
Human approvalEUAI.ART14-03Which agent wrote this code?
Agent, session, and commit, on the pull request.
Change attributionFFIEC.SDLC-02Can the agent log be trusted?
Signed events, append-only, with hashes you can check offline.
Agent event recordsEUAI.ART12-02What if no reviewer is available?
Gated actions are blocked, and the block is recorded.
Approval outage handlingDORA.RES-03
Approval gates require the customer-cloud distribution. The public CLI records activity.
03The record
Every decision, on the record.
Sensitive actions wait for a named reviewer. Decisions export as hash-checked evidence, never raw prompts or source.
Illustrative portal view · demo data

Confirm
A high-risk command waits for a reviewer.
Approve
Tied to the session, command, and reviewer.
Block
A public model call is denied and recorded.
You set the thresholds, own the review, and validate the deployment. Evidence is not a certification. See evidence exports
Connect in one command.
Supported today: Claude Code, Codex, and Cursor. Copilot CLI is roadmap. Setup guide
npm install -g @autodevops/verifier
verifier install --harness claude-code --mode telemetryPrompt text is opt-in.
Built for your cloud.
Where agents run
Developer laptops
Claude Code · Cursor · Codex hooks
CI pipelines
Verifier hooks on pre-commit and pre-push
Your AWS account
- In progress
Portal · ingest · ledger
Signed-event ingest, approvals, append-only audit
- Validated
Worker
IAM-scoped Lambda in private subnets; VPC endpoints, no general egress
- Validated
Amazon Bedrock
Model calls on your own AWS credentials
- Amazon Web Services
- Validated
- Portal stack in your account
- In progress
- Azure and Google Cloud
- Roadmap