FE-564: Flexible turn-response model by lunelson · Pull Request #27 · hashintel/brunch

lunelson · 2026-04-07T15:26:14Z

Stack Context

This stack continues from PR #25 by remodeling choice-turn replies into structured turn responses.
PR #25 establishes the migration and live-turn-card seam; this PR carries the flexible response model above it.

What?

persist choice-turn replies as structured data-turn-response parts with a human-readable summary seam
support zero, one, or many selected options plus optional free-text
keep transcript hydration, resume, and interviewer-context projection aligned with the structured response shape

Why?

The old flow assumed one immediate option selection mapped to one scalar answer string.
This branch retires that assumption so the interview UI can capture richer responses without losing persistence, hydration, or context coherence.
It also formalizes the split from a3c46d1, so the response-model work now lives on its own FE-564 branch and PR.

linear · 2026-04-07T15:26:19Z

FE-564 Flexible turn-response model

lunelson · 2026-04-07T15:26:41Z

This stack of pull requests is managed by Graphite. Learn more about stacking.

lunelson · 2026-04-07T16:18:21Z

Merge activity

Apr 7, 4:18 PM UTC: A user started a stack merge that includes this pull request via Graphite.
Apr 7, 4:24 PM UTC: Graphite rebased this pull request as part of a merge.
Apr 7, 4:24 PM UTC: @lunelson merged this pull request with Graphite.

…-ai/sdk client.messages.stream() (#25) ## Stack Context This stack finishes the SDK migration branch and then layers the flexible turn-response model on top of it. PR #25 is the parent branch for that work; PR #27 carries the response-model changes. ## What? - migrate the interview and observer path onto the current SDK/tooling seam - harden the workspace, controller, rendering, and test boundaries around that seam - project the live turn card before route invalidation so follow-on response-model work has a stable UI boundary ## Why? The FE-564 response-model work needed to be split out on top of `a3c46d1`. That makes this PR the foundation branch again: it now owns the migration, workspace stabilization, and live turn-card projection that the child branch depends on. Updating the metadata keeps the stack legible after the branch split.

augmentcode · 2026-04-07T16:23:03Z

🤖 Augment PR Summary

Summary: This PR remodels choice-turn replies into a structured “turn response” model so interviews can capture zero/one/many option selections plus optional/required free-text.

Changes:

Adds a data-turn-response part schema (selectedOptionIds[] + optional freeText with a “free-text required when no options” rule).
Updates the client turn card to support multi-select (checkboxes), optional additional context, and explicit submit actions for selected vs free-text-only responses.
Updates the /turns/:turnId/select endpoint to accept positions[] and/or freeText, persist both the structured part and a human-readable summary seam, and support many-selections.
Adjusts interviewer context projection to emit response-shaped history when structured response data exists.
Expands regression coverage across client/server tests for persistence, hydration, and projection.
Tooling/infra: introduces oxlint config + typecheck for Vite/Drizzle config; tweaks markdown/highlighting boundaries and observer generation API usage.

_{🤖 Was this summary useful? React with 👍 or 👎}

augmentcode

Review completed. 4 suggestions posted.

Comment augment review to trigger a new review at any time.

augmentcode · 2026-04-07T16:23:04Z

-    };
+    const selectedOptionIds = selectedOptions.map((option) => option.id);
+    const selectedOptionContents = selectedOptions.map((option) => option.content);
+    const responseText = formatTurnResponseText({


formatTurnResponseText() can still return an empty string if selected option content is empty and no freeText is provided; that would persist a data-turn-response without a corresponding text/answer seam. This could break transcript/resume assumptions that every user response yields non-empty visible text.

Severity: medium

_{🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.}

augmentcode · 2026-04-07T16:23:04Z

-  const hasSelection = turn.options.some((o) => o.is_selected);
+  const options = turn.options ?? [];
+  const persistedSelections = options.filter((option) => option.is_selected).map((option) => option.position);
+  const [selectedPositions, setSelectedPositions] = useState<number[]>(persistedSelections);


selectedPositions is initialized from persistedSelections only once; if turn.options later updates (e.g., post-invalidate/hydration) the checkbox state can drift from the persisted selection while inputs are locked by hasPersistedSelection. This can leave the card showing unchecked options even though the server considers the turn answered.

Severity: low

_{🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.}

augmentcode · 2026-04-07T16:23:04Z

      lines.push(questionLine);
    }
-    if (turn.answer) lines.push(`Answer: ${turn.answer}`);
+    const selectedOptions =


buildInterviewerContext() derives chosen options from option.is_selected but derives free-text from persisted data-turn-response; if those two sources ever diverge, interviewer context could become internally inconsistent with the canonical persisted response. Consider projecting both chosen options and free-text from the same persisted response record when present.

Severity: medium

_{🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.}

augmentcode · 2026-04-07T16:23:04Z

+      }
+      const uniquePositions = [...new Set(positions)];
+      const selectedOptions = findTurnOptionsByPositions(turn, uniquePositions);
+      if (selectedOptions.length !== uniquePositions.length) {


If positions contains anything not found in turn.options, submitTurnResponse() returns early without surfacing an error, which could make stale-state cases look like a no-op to users. Consider ensuring this path reports a failure through the existing mutation error surface so the UI can explain what happened.

Severity: low

_{🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.}

This was referenced Apr 7, 2026

FE-538: Entity sidebar (read-only) #24

Merged

FE-559: Migrate: @anthropic-ai/claude-agent-sdk query() -> @anthropic-ai/sdk client.messages.stream() #25

Merged

lunelson changed the title ~~chore: restore lint baseline and fix deprecated APIs~~ FE-564: Flexible turn-response model Apr 7, 2026

lunelson marked this pull request as ready for review April 7, 2026 16:17

lunelson mentioned this pull request Apr 7, 2026

FE-566: Generic knowledge layer schema and sidebar projection #28

Merged

lunelson changed the base branch from ln/fe-559-migrate-sdk to graphite-base/27 April 7, 2026 16:21

lunelson changed the base branch from graphite-base/27 to main April 7, 2026 16:22

augmentcode Bot reviewed Apr 7, 2026

View reviewed changes

lunelson added 5 commits April 7, 2026 16:23

chore: restore lint baseline and fix deprecated APIs

bd6d385

chore: expand tooling coverage and fix config warnings

2f18f6f

chore: trim rich markdown rendering features

7de0d9f

docs: record trimmed rich markdown scope

48b8ae7

feat: persist single-option turn responses with optional free-text

a0cbd35

lunelson added 2 commits April 7, 2026 16:23

feat: support free-text-only turn responses

79a7b5a

feat: support many-selection turn responses

67e3a59

lunelson force-pushed the ln/fe-564-flexible-turn-response-model branch from 3852b56 to 67e3a59 Compare April 7, 2026 16:23

lunelson merged commit f6570c2 into main Apr 7, 2026
2 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

FE-564: Flexible turn-response model#27

FE-564: Flexible turn-response model#27
lunelson merged 7 commits into
mainfrom
ln/fe-564-flexible-turn-response-model

lunelson commented Apr 7, 2026 •

edited

Loading

Uh oh!

linear Bot commented Apr 7, 2026

Uh oh!

lunelson commented Apr 7, 2026 •

edited

Loading

Uh oh!

lunelson commented Apr 7, 2026 •

edited

Loading

Uh oh!

augmentcode Bot commented Apr 7, 2026

Uh oh!

augmentcode Bot left a comment

Uh oh!

augmentcode Bot Apr 7, 2026

Uh oh!

augmentcode Bot Apr 7, 2026

Uh oh!

augmentcode Bot Apr 7, 2026

Uh oh!

augmentcode Bot Apr 7, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

lunelson commented Apr 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Stack Context

What?

Why?

Uh oh!

linear Bot commented Apr 7, 2026

Uh oh!

lunelson commented Apr 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lunelson commented Apr 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merge activity

Uh oh!

augmentcode Bot commented Apr 7, 2026

Uh oh!

augmentcode Bot left a comment

Choose a reason for hiding this comment

Uh oh!

augmentcode Bot Apr 7, 2026

Choose a reason for hiding this comment

Uh oh!

augmentcode Bot Apr 7, 2026

Choose a reason for hiding this comment

Uh oh!

augmentcode Bot Apr 7, 2026

Choose a reason for hiding this comment

Uh oh!

augmentcode Bot Apr 7, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

lunelson commented Apr 7, 2026 •

edited

Loading

lunelson commented Apr 7, 2026 •

edited

Loading

lunelson commented Apr 7, 2026 •

edited

Loading