Flock Synthetic CLI
Run synthetic users from terminal workflows and CI jobs with deterministic artifacts.
Artifacts and outputs
- output.json
- report.md
- engineering_events.json
Launch-path documentation
Developer surfaces for Flock
These docs describe the CLI and GitHub App surfaces we are shaping for teams that want to run synthetic users, inspect findings, and act on evidence.
Each surface follows the same standard: findings should be grounded, reviewable, and useful to the developers or agents fixing the issue. We continuously refine the target shape from customer feedback.
Flock GitHub App
Planned PR checks will run synthetic journeys and post evidence-backed findings to GitHub.
Artifacts and outputs
- Inline PR comments
- PR summary
- Check run status
Quality bar
Evidence-backed or it does not ship
Every claim needs evidence
Findings must tie to observable artifacts: UI text, screenshots, element states, console/runtime errors, or network traces.
Every issue should be addressable
We optimize for findings that map to concrete code and product changes teams can address in a pull request.
Hallucination resistance by design
Agents are constrained to evidential signals and structured outputs to reduce speculative or non-actionable feedback.