kandev/docs/plans/initial-task-brief/task-03-transcript-evidence.md

id, title, status, wave, depends_on, plan, requirements, acceptance_criteria, system_design

id	title	status	wave	depends_on	plan	requirements	acceptance_criteria	system_design
03-transcript-evidence	Transcript visibility evidence	done	3	02-first-message-composition	plan.md	REQ-TASKS-INITIAL-TASK-BRIEF-001	AC-TASKS-INITIAL-TASK-BRIEF-001.1 AC-TASKS-INITIAL-TASK-BRIEF-001.2 AC-TASKS-INITIAL-TASK-BRIEF-001.3 AC-TASKS-INITIAL-TASK-BRIEF-001.7	../../specs/tasks/system-design/initial-task-brief.md

Task 03: Transcript visibility evidence

Summary

Prove that the stored first prompt replaces the synthetic brief without losing content. Document the behavior only after the implementation passes its focused checks.

In scope

• Extend the processed-message regression with a combined stored prompt #1.
• Add desktop and phone composer/reload flows using mock-agent fixtures.
• Add a concise explanation to the existing public task guide.
• Record actual command results in all work orders and the manifest.

Out of scope

Do not change unrelated launch policy, historical messages, or public API fields.

Acceptance

• Both desktop and phone preserve brief plus instruction after submission and reload.
• The synthetic row disappears, prompt #1 exists once, and later messages do not repeat the brief.
• Long content remains reachable with existing controls and no document-level horizontal overflow.

ASCII UI preview

Same UI-01 as the package preview, covering AC-TASKS-INITIAL-TASK-BRIEF-001.2 and .7.

UI-01: Chat after first message (desktop and phone)

Before                         Proposed
+-------------------------+    +-------------------------+
| User #1                 |    | User #1                 |
| rebase on latest main   |    | Original task brief     |
|                         |    |                         |
| Agent response          |    | rebase on latest main   |
+-------------------------+    | Agent response          |
| Composer         [Send] |    +-------------------------+
+-------------------------+    | Composer         [Send] |
                               +-------------------------+

Both viewports retain their existing Chat surface and vertical scroll owner. Phone controls keep the existing safe-area and touch behavior. No new controls are planned.

Verification

Start with failing behavior assertions before implementation. Then run this block from the repository root:

(cd apps && pnpm install --frozen-lockfile)
(cd apps/web && pnpm exec vitest run hooks/use-processed-messages-fallback.test.ts)
(cd apps/web && pnpm e2e:run --project chromium tests/chat/initial-task-brief.spec.ts)
(cd apps/web && pnpm e2e:run --project mobile-chrome tests/chat/mobile-initial-task-brief.spec.ts)
python3 scripts/lint-spec-files.py --all
git diff --check

Files likely touched

• apps/web/hooks/use-processed-messages-fallback.test.ts
• apps/web/e2e/tests/chat/initial-task-brief.spec.ts (new)
• apps/web/e2e/tests/chat/mobile-initial-task-brief.spec.ts (new)
• apps/web/e2e/tests/chat/initial-task-brief-helpers.ts (new, if shared)
• docs/public/tasks-and-workflows.md
• docs/plans/initial-task-brief/plan.md

Dependencies

Task 02 must be complete.

Risks

Use fresh managed builds and causal waits. A seeded API assertion alone is not browser evidence. Do not weaken existing transcript pagination or fallback eligibility to satisfy these tests.

Parallelism

sequential

Inputs

• Requirements.
• System design.
• Package evidence and test matrix.

Results

• Extended the processed-message regression so a persisted combined prompt replaces the synthetic task-description row without duplicating the brief.
• Added managed desktop and phone Chat flows that submit the first message, reload, verify the stored prompt, send a later message, and check for document overflow.
• Added the same prepared-session API option to both E2E flows without changing production API behavior.
• Added public guidance to docs/public/tasks-and-workflows.md.
• pnpm exec vitest run hooks/use-processed-messages-fallback.test.ts passed.
• pnpm e2e:run --project chromium tests/chat/initial-task-brief.spec.ts passed.
• pnpm e2e:run --project mobile-chrome tests/chat/mobile-initial-task-brief.spec.ts passed.