Every file read, every command run, every claim verified — four independent analyzers surface the truth in seconds.
Built for teams shipping AI-assisted code
Trust Score
Weighted across 4 analyzers
Honesty Check
Mismatch detected
3 discrepancies
Ripple Risk
High — 14 deps
auth.ts touched
QA Lens
2 fake assertions
expect(true).toBe(true)
Timeline
23 operations
4 rewrites detected
Claimed: "Added auth middleware and updated tests"·8 files modified, 2 test files deleted
Supported agents
Agent format auto-detected. No config, no SDK, no code changes.
Anthropic's CLI agent. Full parsing of JSONL conversation logs — tool calls, file edits, bash commands.
Trusted by engineering teams
“Caught our agent deleting test files and calling it a refactor. Videlic literally saved our deploy.”
Marcus T.
Staff Engineer · Series B startup
“The Honesty Check is wild. Our agent claimed it 'updated tests' — Videlic showed it just removed them.”
Priya K.
Engineering Lead · DevTools company
“We run Videlic on every PR now. It's the first thing we check before merging agent-written code.”
Daniel R.
Senior Developer · Fintech team
How it works
Pipe your agent's session log via CLI, API, or CI/CD integration. One command — works with any supported agent.
Honesty check, ripple risk, QA lens, and timeline extraction run in parallel. Partial failures never block results.
A permanent, shareable URL with trust score, discrepancies, and concrete evidence your team can act on.
“We run Videlic on every PR. It’s the first thing we check before merging agent code.”— Senior Developer, Fintech
Your agent wrote the code. Videlic tells you what actually happened. Free plan, no credit card.