feat: exhibit floor design — revive the RepoVis vision as the shipped UI

[Coldaine]

User description

What

Design doc for The Exhibit Floor: restoring the repo`s original purpose — an agent with a rich prompt authoring interesting artifacts, rendered by exhibit-grade UI we build in advance.

The RepoVis vision (5-view working prototype + ~40 exhibit concepts) entered the repo 2026-04-05 (3ce52d8) and was shelved to docs/ideas/repoviz/ on 2026-05-30 without ever shipping. docs/north-star.md calls it "the bar the Render pillar is aiming for." This plan makes the app hit the bar:

• Exhibit vocabulary v1: 7 typed components — the prototypes five views (relationship DAG, folding activity narrative, walkthrough, concern snapshot, momentum) + seismograph + thermal map from the idea banks best creative directions, re-aimed at agent-session observation.
• Artifact envelope with mandatory narrative, relevance, decay, and status — the model curates a gallery, it doesn`t fill dashboard slots.
• Curator prompt v2: big packets, gallery state fed back as memory, output = gallery ops (create/refresh/retire).
• ExhibitStage: museum-floor surface with idle curatorial cycling; the existing canvas graph becomes one exhibit (live_graph) in the rotation.
• Validation: replay the homelab Codex corpus live and dump the gallery timeline against human ground truth.

Stacked on #114 (replay runner). Implementation follows as PR-B (components + stage), PR-C (curator pipeline), PR-D (corpus gallery run).

🤖 Generated with Claude Code

https://claude.ai/code/session_01U9EpjZNoWaXLzfG93tdvTU

---

CodeAnt-AI Description

Lay out the Exhibit Floor plan for the new gallery-style UI

What Changed

• Adds a design plan for replacing the fixed dashboard with a curated “exhibit floor” that shows agent work as a rotating gallery of typed artifacts
• Defines the seven exhibit types, the required narrative-focused artifact format, and the curator behavior that creates, refreshes, or retires exhibits based on relevance
• Describes the new stage layout, including the gallery list, active exhibit view, commentary panel, and archive shelf, plus the planned replay-based validation flow

Impact

✅ Clearer product direction
✅ Stronger support for gallery-style agent sessions
✅ Fewer dashboard-first assumptions

💡 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]

🤖 CodeAnt AI — Review Status

Status	Commit	Started (UTC)	Finished (UTC)
✅ Reviewed your PR	3203ee9	Jul 24, 2026 · 01:53	01:54

[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

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 2650b3ae-0425-40bd-88bf-46afe4c19a50

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feat/exhibit-floor

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[Copilot]

Pull request overview

Adds a new design plan document that defines “The Exhibit Floor” direction: shifting the app from a fixed dashboard to a curated, rotating gallery of typed exhibits authored by the shadow model, aligned with the shelved RepoVis prototype and idea bank.

Changes:

• Introduces the Exhibit Floor concept and “curator” inversion (prebuilt exhibit vocabulary + model-authored narrative artifacts).
• Specifies the v1 exhibit vocabulary, artifact envelope, curator prompt v2 behavior, and ExhibitStage surface layout.
• Describes replay-based validation criteria and a PR series breakdown (PR-A through PR-D).

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[charliecreates]

Review

Blocking findings

• plan-exhibit-floor.md:83-99,151-152 and curator-prompt-v2.md:23-31,169-171 — the packet contract cannot fit the current inference pipeline. The plan targets 60–100k-token packets containing full recent detail, an earlier digest, and the gallery, while docs/domain-inference.md:77-88 defines an approximately 10.5k-token budget and src/inference/prompt-builder.ts:21-30,68-99 currently builds only bounded recent context. The plan also makes the 100k-token packager overhaul a non-goal and says PR-C keeps the current packager. Please make the packager overhaul a prerequisite/part of PR-C with explicit truncation and size limits, or revise the curator contract to a budget the existing pipeline can honor; otherwise “the gallery is the memory” is not implementable as written.

• plan-exhibit-floor.md:61-72,87-89 versus curator-prompt-v2.md:44-48,56-69 — the gallery wire contract is internally inconsistent. The plan models retire as an artifact operation and includes status: 'retired'; the prompt models retire as { artifactId, reason } and allows only fresh | active | stale artifact status. Choose one canonical galleryOps/envelope contract before PR-C, including where the retirement reason and archived state live.

• plan-exhibit-floor.md:93-94 versus curator-prompt-v2.md:38-48,172-175 — the backward-compatibility pulse is incomplete. The existing parser/renderer contract still has predictedNextAction/predictedNextConfidence, per-signal risks, and attention.intent (src/inference/response-parser.ts:17-26,177-197), but the new pulse contains only phase, phase confidence, risk level, and headline. The notes promise old next_move, risk, and objective insights from pulse + galleryOps without defining a deterministic mapping, especially when galleryOps is empty. Add the missing pulse fields or specify the exact fallback/mapping so the ghost trail, risk surface, and objective do not silently regress.

Non-blocking follow-up

• plan-exhibit-floor.md:123-130 — the validation fixture is not present. tests/fixtures/transcripts/codex/rollout-2026-07-23-homelab-coordinator.jsonl does not exist on this base/head, and the PR adds only two design documents. Add or identify the fixture and ground-truth source, or explicitly make fixture capture part of PR-D and adjust the acceptance wording.

• curator-prompt-v2.md:3-6 — avoid creating a second prompt source of truth. AGENTS.md:24,42-47 and docs/architecture.md:69-74 require src/inference/prompts.ts to be the single authoritative prompt. Keep this document explicitly non-authoritative review material, or move the canonical text/rationale into src/inference/prompts.ts when PR-C lands so the copies cannot drift.