Skip to content

docs(coverage): expose inland downloads before hydration - #476

Draft
dodeja wants to merge 2 commits into
mainfrom
codex/coverage-static-directory-downloads
Draft

dodeja wants to merge 2 commits into
mainfrom
codex/coverage-static-directory-downloads

Conversation

@dodeja

@dodeja dodeja commented Oct 19, 2026 •

Copy link
Copy Markdown
Member

Why

The weekly published-coverage check cannot find inland rail JSON/CSV anchors in the initial HTML. The files and complete Markdown references are available, but the human download links live inside the interactive directory component.

What changed

  • Move JSON, CSV and complete-table links into static human-visible page Markdown.
  • Remove the client-component copy to avoid duplicate links after hydration.
  • Check that regenerated human content retains the public download links independently of the component.

Scope

Inland-directory link discoverability only. Coverage values, identifiers, dates, filtering and agent table content remain unchanged. Automatic data refresh remains blocked on a validated complete source export, retrieval and normalization.

Blast Radius

The human download links move below the directory footer. No API, SDK or gateway code changes. This draft does not deploy anything; actual hosted HTML anchor presence needs verification after publication.

Verification

  • Read-only live verification passed 20 of 21 checks: all eight downloads, eight complete Markdown references and the agent index passed. Inland HTML lacked native download anchors; an independent HTML parser confirmed the gap.
  • The tightened human-section regression failed before the move and passes afterward. All 45 coverage tests and both generator drift checks pass. Focused lint and whitespace checks pass.
  • Independent source review approved static-link ownership and regeneration preservation. No fresh source data was collected and the published data date was preserved.

View with [code]smith Autofix with [code]smith
Need help on this PR? Tag @codesmith-bot with what you need.

@vercel

vercel Bot commented Oct 19, 2026 •

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated
api Ready Ready Preview Oct 19, 2026 4:03pm UTC

Request Review

@mintlify

mintlify Bot commented Oct 19, 2026 •

Copy link
Copy Markdown
Contributor

Preview deployment for your docs. Learn more about Mintlify Previews.

Project Status Preview Updated
terminal49 🟢 Ready View Preview Oct 19, 2026, 4:03 PM

@github-actions github-actions Bot added the risk-shadow:high Shadow-only risk signal: high label Oct 19, 2026
@github-actions

Copy link
Copy Markdown
Contributor

Shadow risk calibration

Signal Result Authority
Path classifier risk:medium Deterministic advisory policy
heuristic fallback risk-shadow:high Shadow only

Calibration result: disagreement (medium vs high).

Shadow risk never blocks or enables merge. Keep it in observation mode until roughly 50 pull requests have human-reviewed labels, then calibrate before considering any policy use.

@github-actions github-actions Bot added the risk:medium Advisory path risk; current-head human review needed label Oct 19, 2026
@github-actions github-actions Bot added risk:unknown Advisory evidence unavailable or changed; no merge permission risk:medium Advisory path risk; current-head human review needed and removed risk:medium Advisory path risk; current-head human review needed risk:unknown Advisory evidence unavailable or changed; no merge permission labels Oct 19, 2026

This branch was successfully deployed

2 active deployments
staging - docs — 261ed97c Deployed Oct 19, 2026 by mintlify[bot]
Preview — 261ed97c Deployed Oct 19, 2026 by vercel[bot]
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

risk:medium Advisory path risk; current-head human review needed risk-shadow:high Shadow-only risk signal: high

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant