Skip to content

Refresh machine-readable website vocabulary facts through staging #76

Description

@hwesterb

Problem

The canonical vocabulary now contains 453 patterns, but web/public/llms.txt and the checked-in built src/sema/server/static/llms.txt still advertise 452. These files were deliberately left out of PR #74 because website changes must use the separate Railway staging service and receive explicit production approval.

Proposed approach

Prefer deriving the count during the website build, or remove the exact count when it adds no value, so future vocabulary additions cannot stale the machine-readable site description. Rebuild static assets only through the existing frontend workflow.

Acceptance criteria

  • Source and built llms.txt agree with the canonical vocabulary, without a manually maintained count if practical.
  • Deploy the candidate commit to semahash-staging with the workspace/production flags unchanged.
  • Verify the staged llms.txt and normal public pages.
  • Do not promote to production until Henrik explicitly approves the website change.

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

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions