fix(ui): keep analysis-run next actions in the list name (v0.87.1)

[cursor]

Successor to dirty #149 after #148 (kind-specific pending copy) and #145 (ADR 0020 retention purge / Storybook tokens, v0.87.0) landed on feat/role-responsibility-agent-ontology (69c035b). This slice is AccName only (v0.87.1).

Buyer gap after #148: list aria-label still hid the next-action sentence, so assistive technology only heard kind · status · entity.

• List accessible name is Open analysis run: {caption}. {nextAction} when a next action exists (WCAG 2.2 SC 4.1.2 / AccName 1.1)
• Detail still repeats the same sentence
• Pending TEPP corpus says measurement has not started — it does not say the run is already finishing
• Running TEPP keeps the in-flight corpus sentence
• Unknown kind/status defaults do not echo a raw wire code
• Folded into ADR 0014. Do not add another ADR 0020 (feat: add granted retention purge and Storybook tokens (v0.87.0) #145 owns that number)

Do not merge before #74. Do not land dirty #149, stale #146, or draft #163. Do not mix #142 start reconstruction, #153 person catalog, #157 lineage-only create, #161 seeded period-report registry, or a live TEPP transport. Do not invent a theta.

Next action: Open a Failed TEPP row and hear the connect-measurement sentence in the list name. Request a lineage reconstruction: the Pending lineage name still says reconstruction. A Pending TEPP fixture must not say Reconstruction and must not say the run has already started finishing.

Verified: frontend 62 tests + oxlint + production build after merge onto #145.

[cursor]

Stale comment

SOUND for the AccName slice on cd5c0af (v0.87.1).

Next action: Land this onto feat/role-responsibility-agent-ontology after Checks are green. Do not merge to main before #74. Do not land dirty #149 / #146 (they steal ADR 0020 / v0.86.x), draft #165 (subset), #163 (pre-0.87.0), or #138 (main). Do not take ADR 0020 here — #145 already owns it.

What this head does:

• List aria-label is Open analysis run: {caption}. {nextAction} when a next action exists (WCAG 2.2 SC 4.1.2 / AccName 1.1).
• Pending TEPP AccName and corpus say measurement has not started and is not a calibrated result. They do not say Reconstruction and do not say the run is already finishing.
• Failed lineage / TEPP / period-report names stay kind-specific.
• Unknown kind/status defaults do not echo a raw wire code.
• Folded into ADR 0014. Package is 0.87.1.

Residual, not a blocker on this slice:

• Running and cancelled TEPP list names stay caption-only (nextAction is null). Open those rows to hear the in-flight / cancelled corpus sentence.
• #169 also claims v0.87.1 for Storybook home tokens — retarget that package number if both land.
• Do not mix #142/#167 start reconstruction, #153 person catalog, #157/#164 create-gate, or a live TEPP transport into this PR.

Verified by reading analysisRunAccessibleName / pending-TEPP tests on this head. Frontend Checks were still queued at review time — treat a red Frontend job as the merge gate, not this comment.

  

_{Sent by Cursor Automation: Fix Issues}

[cursor]

Stale comment

SOUND on cd5c0af for the analysis-run AccName slice (v0.87.1).

Next action: Prefer successor #173 (9d9990a, v0.87.2) if both are open. That branch already contains this rebase plus calendar / period-report / post-list AccName (ADR 0024). Land one stack, not both.

Do not merge to main before #74. Do not land dirty #149 / #146, draft #165, #163, or #138. Do not take ADR 0020. Do not mix start reconstruction, person catalog, create-gate, Storybook tokens, or a live TEPP transport.

Residual on this head only: running and cancelled TEPP list names stay caption-only. Open those rows for the in-flight / cancelled corpus sentence.

View PR

  

_{Sent by Cursor Automation: Fix Issues}

[cursor]

Verdict: COMMENT — do not merge this draft

Same AccName slice as #163. The landing vehicle is #188 45a5992: running and cancelled TEPP list names include the next-action sentence, Storybook inventories AnalysisRunListButton, and AGENTS.md holds the AccName formula.

Next action: Close this draft in favor of #188. Open a Running TEPP fixture there and hear that measurement is not a calibrated result.

View PR

  

_{Sent by Cursor Automation: Fix Issues}