feat: UX hardening P1/P2 — Copilot/Voice, readability, Consistency Checker, AI context

[qnbs]

User description

Context

Part 2 of the user-facing UX/interaction-risk hardening set. Part 1 (#191, P0) is merged; this PR carries the P1/P2 items and all CodeAnt-review-wave fixes, rebased directly onto main.

Two audit claims were stale and are not changed here: the Dashboard AI logline is already a safe preview flow (only the readability rendering was a real gap).

Items

3 · Copilot & Voice spatial separation (P1)

The Copilot FAB and Voice panel both anchored at bottom-20/right-4 and overlapped when both flags were on. The voice panel now stacks above the FAB (bottom-[9.25rem] / md:bottom-[5.25rem]), clearing the FAB and the mobile bottom-nav. Adds a visible mode chip (Commands/Dictation).

4 · Dashboard readability scale + logline undo hint (P1)

Readability score now shows /100 plus an interpretation Tooltip (Flesch 0–100, higher = easier). Applying an AI logline fires an undo-hint toast (project slice is redux-undo wrapped).

5 · Consistency Checker structured output (P1)

The consistencyCheck prompt now requests a JSON finding array; the hook parses it into a discriminated ConsistencyResult (robust raw-text fallback so the pane never regresses to empty), and the view renders a severity-tagged list instead of a raw <pre>. CodeAnt waves hardened this further: parse-drop arrays surface as text (no false all-clear), a shape-based isAbortError guard preserves results on cancellation, and the empty-array clean-state contract is honoured end-to-end.

6 · AI context-source labels (P2)

Context notes so users see what feeds the AI: SceneBoard beat modal ("scene titles + opening lines, ~2,000 chars") and the Copilot panel header ("current view + project metadata").

Process docs

docs(process): documents the >100-file PR-split rule (CodeAnt skips larger PRs) as runbook §1a, and the plan-mode parallel-exploration exception to the low-end-hardware shell rule.

Tests & i18n

• Unit tests across voice mode chip, dashboard undo toast, and consistency JSON parse / text-fallback / abort / parse-drop cases.
• i18n: voice/dashboard/consistency/sceneboard/copilot keys across all 17 locales (including proper translations for the 12 non-core locales) + rebuilt bundles.
• Local gate green: suppressions OK, lint, typecheck, i18n:check, targeted vitest.

🤖 Generated with Claude Code

---

CodeAnt-AI Description

Show consistency checks as structured findings and clarify AI-facing UI

What Changed

• Consistency checks now show a clear list of findings with severity labels, and an empty result now displays a plain “no findings” state instead of raw output.
• Aborted consistency checks no longer clear a previous result, including cancellations that arrive as plain abort-shaped objects.
• The Copilot and voice panels now explain what data they use, and the voice panel shows a visible mode chip for commands vs dictation.
• Dashboard readability now shows /100 with an explanation tooltip, and applying an AI logline now shows an undo hint.
• Tooltips now open for keyboard users when the trigger is focused, not only on hover.

Impact

✅ Clearer consistency check results
✅ Fewer lost results after canceled AI checks
✅ More obvious AI context and voice mode labelstitle: |-
Show structured consistency results, voice mode labels, and clearer AI hints

💡 Usage Guide

Checking Your Pull Request

Every time you make a pull request, our system automatically looks through it. We check for security issues, mistakes in how you're setting up your infrastructure, and common code problems. We do this to make sure your changes are solid and won't cause any trouble later.

Talking to CodeAnt AI

Got a question or need a hand with something in your pull request? You can easily get in touch with CodeAnt AI right here. Just type the following in a comment on your pull request, and replace "Your question here" with whatever you want to ask:

@codeant-ai ask: Your question here

This lets you have a chat with CodeAnt AI about your pull request, making it easier to understand and improve your code.

Example

@codeant-ai ask: Can you suggest a safer alternative to storing this secret?

Preserve Org Learnings with CodeAnt

You can record team preferences so CodeAnt AI applies them in future reviews. Reply directly to the specific CodeAnt AI suggestion (in the same thread) and replace "Your feedback here" with your input:

@codeant-ai: Your feedback here

This helps CodeAnt AI learn and adapt to your team's coding style and standards.

Example

@codeant-ai: Do not flag unused imports.

Retrigger review

Ask CodeAnt AI to review the PR again, by typing:

@codeant-ai: review

Check Your Repository Health

To analyze the health of your code repository, visit our dashboard at https://app.codeant.ai. This tool helps you identify potential issues and areas for improvement in your codebase, ensuring your repository maintains high standards of code health.

[codeant-ai]

---

Thanks for using CodeAnt! 🎉

We're free for open-source projects. if you're enjoying it, help us grow by sharing.

Share on X ·
Reddit ·
LinkedIn

[vercel]

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

Project	Deployment	Actions	Updated (UTC)
worldscript-studio	Ready	Preview, Comment	Jun 20, 2026 10:19pm

[qnbs]

@CodeAnt-AI review

[codeant-ai]

---

Thanks for using CodeAnt! 🎉

We're free for open-source projects. if you're enjoying it, help us grow by sharing.

Share on X ·
Reddit ·
LinkedIn

[codeant-ai]

Sequence Diagram

This PR changes the character consistency checker so that AI responses are parsed into structured severity-tagged findings, with a robust fallback to raw text when JSON output is missing or malformed.

sequenceDiagram
    participant User
    participant ConsistencyView
    participant ConsistencyHook
    participant GeminiService
    participant AIProvider

    User->>ConsistencyView: Click Check Consistency for character
    ConsistencyView->>ConsistencyHook: runCheck(characterId)
    ConsistencyHook->>GeminiService: Build consistencyCheck prompt from context
    GeminiService-->>ConsistencyHook: Prompt text
    ConsistencyHook->>AIProvider: generateText(prompt, options, abortSignal)
    AIProvider-->>ConsistencyHook: AI response text

    alt AI returns JSON array of findings
        ConsistencyHook->>ConsistencyHook: parseConsistencyResult to structured findings
        ConsistencyHook-->>ConsistencyView: checkResult = structured findings
        ConsistencyView-->>User: Render severity-tagged findings list
    else AI returns malformed JSON or plain text
        ConsistencyHook->>ConsistencyHook: parseConsistencyResult to text fallback
        ConsistencyHook-->>ConsistencyView: checkResult = text result
        ConsistencyView-->>User: Render raw text block (never empty)
    end

Loading

---

Generated by CodeAnt AI

[qnbs]

@CodeAnt-AI review

[codeant-ai]

---

Thanks for using CodeAnt! 🎉

We're free for open-source projects. if you're enjoying it, help us grow by sharing.

Share on X ·
Reddit ·
LinkedIn

[codeant-ai]

Sequence Diagram

This diagram shows how the consistency checker now requests structured JSON findings from the AI and preserves results on cancellation, and how applying an AI logline on the dashboard updates state while showing an undo-hint toast.

sequenceDiagram
    participant User
    participant ConsistencyChecker
    participant AIService
    participant Dashboard
    participant ToastService

    User->>ConsistencyChecker: Click Check Consistency for character
    ConsistencyChecker->>AIService: Request consistency check with JSON findings prompt
    AIService-->>ConsistencyChecker: Return JSON array or plain text
    ConsistencyChecker->>ConsistencyChecker: Parse into ConsistencyResult (structured or text)
    ConsistencyChecker-->>User: Show severity list, clean state, or text fallback

    alt Later consistency check is aborted
        User->>ConsistencyChecker: Trigger new check then close panel or navigate away
        AIService-->>ConsistencyChecker: AbortError cancellation
        ConsistencyChecker-->>User: Stop checking but keep previous results
    end

    User->>Dashboard: Apply AI logline suggestion
    Dashboard->>Dashboard: Update project logline state (undoable)
    Dashboard->>ToastService: Show success toast with undo hint
    ToastService-->>User: Display logline updated with Ctrl+Z message

Loading

---

Generated by CodeAnt AI

[qnbs]

@CodeAnt-AI review

[codeant-ai]

---

Thanks for using CodeAnt! 🎉

We're free for open-source projects. if you're enjoying it, help us grow by sharing.

Share on X ·
Reddit ·
LinkedIn

[codeant-ai]

Sequence Diagram

This diagram shows how the Consistency Checker now calls the AI provider for a JSON findings array, parses it into structured results for display, and uses a shared abort guard so cancelled runs do not clear existing findings.

sequenceDiagram
    participant User
    participant Dashboard
    participant ConsistencyChecker
    participant AIProvider
    participant AIModel

    User->>Dashboard: Click Check Consistency for character
    Dashboard->>ConsistencyChecker: runCheck(characterId)
    ConsistencyChecker->>AIProvider: Request consistency check text
    AIProvider->>AIModel: Send prompt with JSON findings contract
    AIModel-->>AIProvider: Return findings array or text
    AIProvider-->>ConsistencyChecker: Deliver response or abort signal

    alt Valid JSON findings
        ConsistencyChecker->>ConsistencyChecker: parseConsistencyResult to structured findings
        ConsistencyChecker-->>Dashboard: checkResult structured (findings or empty)
        Dashboard-->>User: Show severity list or no findings message
    else Cancelled by user
        ConsistencyChecker->>ConsistencyChecker: Detect abort with isAbortError and keep prior result
        ConsistencyChecker-->>Dashboard: Stop checking without changing checkResult
    else Invalid or malformed output
        ConsistencyChecker->>ConsistencyChecker: Fall back to text result
        ConsistencyChecker-->>Dashboard: checkResult as raw text
        Dashboard-->>User: Show plain text block
    end

Loading

---

Generated by CodeAnt AI