Skip to content

[Evidence Sprint follow-up] Define live-pilot entry criteria for external claims #3586

Description

@ootakazuhiko

Summary

Define live-pilot entry criteria before converting fixture-backed Evidence Sprint pilot materials into external or live-behavior claims.

Background

Issues #3572 and #3573 produced fixture-backed/report-only Web API and event-driven pilot evidence. The release assets in #3575 intentionally avoid live API/event behavior, productivity, safety, quality, adoption-impact, and vendor-comparison claims. Before any public external pilot claim, the project should define what evidence, consent, baselines, and operational boundaries are required.

Proposed scope

  • Define minimum entry criteria for live/external pilots, including repository consent, data handling, artifact retention, and claim boundaries.
  • Specify which Evidence Sprint metrics must be collected before a live-pilot claim can be made.
  • Separate pilot readiness criteria from enforcement or benchmark claims.
  • Link the criteria from the measurement protocol or pilot evidence docs as appropriate.

Acceptance criteria

  • Live-pilot entry criteria are documented with explicit supported/unsupported claim boundaries.
  • Required evidence artifacts and measurement fields are listed.
  • External repository, data-retention, and consent assumptions are explicit.
  • Fixture-backed pilots remain clearly marked report-only until the criteria are satisfied.

Evidence references

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    area:evidenceEvidence and artifactscodexTask intended to be executed through Codex CLIpriority:p2Later or follow-up slicetype:taskImplementation task

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions