Skip to content

replay: first live curator gallery run over the homelab corpus#118

Open
Coldaine wants to merge 1 commit into
feat/curator-pipelinefrom
feat/exhibit-gallery-run
Open

replay: first live curator gallery run over the homelab corpus#118
Coldaine wants to merge 1 commit into
feat/curator-pipelinefrom
feat/exhibit-gallery-run

Conversation

@Coldaine

@Coldaine Coldaine commented Jul 24, 2026

Copy link
Copy Markdown
Owner

User description

What

PR-D of the exhibit floor series (#115 design -> #116 components -> #117 pipeline -> this): the first end-to-end proof that an agent with a rich prompt authors interesting artifacts against UI built in advance.

Replayed the homelab PG18-recovery corpus (2,885 events, 4.6 virtual hours) at 60x with the curator prompt live on DeepSeek:

  • 125 triggers -> 16 curator calls (the raised 25-event/120s floor doing its job); p50 19.9s / p95 32.9s.
  • The curator authored 17 exhibits across six of the seven types, retiring 14 as the story moved on - every retirement carries a visitor-facing reason (e.g. "a stunning but completed satellite; the session`s center of gravity has moved to PR merge mechanics").
  • Final floor is honest about the unfinished ending: momentum "Two blockers, three hopes" (gauge 48; PRS MERGED 0; DRY-RUN ATTEMPTS FAILED 3), a fresh concern snapshot "The dry-run validation wall", and walkthrough "Package rebrand breaks stalemate".
  • Ground-truth checkpoints: scope-drift abort, plan pivot, and GHCR-403 stall surfaced; the final-not-done checkpoint was missed by the text matcher even though the momentum exhibit plainly reads as not-done - the matcher-precision follow-up stands.
  • Gallery + full report archived under docs/history/gallery-run-2026-07-23/; findings appended to docs/history/log.md.

Known UI nit (follow-up)

Momentum what`s-next confidence chips render "0.85%" instead of "85%" - 0..1 vs 0..100 scale mismatch between the curator contract and the chip formatter.

🤖 Generated with Claude Code

https://claude.ai/code/session_01U9EpjZNoWaXLzfG93tdvTU


CodeAnt-AI Description

Record the first live curator gallery run over the homelab corpus

What Changed

  • Adds the archived gallery and report for the first end-to-end live curator run over the 2,885-event homelab recovery corpus
  • Records the run outcome in the history log, including trigger count, curator call count, gallery size, and the final active exhibits
  • Captures the run’s notable checkpoints, including the dry-run validation stall, the package rebrand that unblocked publishing, and the final not-done state

Impact

✅ Clearer run history
✅ Easier audit of curator outcomes
✅ Faster review of gallery results

💡 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.

16 DeepSeek curator calls over the 2,885-event corpus produced 17 exhibits
(3 active at session end, 14 retired with visitor-facing reasons). Gallery
and report archived; findings appended to the history log.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01U9EpjZNoWaXLzfG93tdvTU
Copilot AI review requested due to automatic review settings July 24, 2026 03:37
@codeant-ai

codeant-ai Bot commented Jul 24, 2026

Copy link
Copy Markdown

🤖 CodeAnt AI — Review Status

Status Commit Started (UTC) Finished (UTC)
✅ Reviewed your PR 7c2ad40 Jul 24, 2026 · 03:37 03:37

@codeant-ai

codeant-ai Bot commented Jul 24, 2026

Copy link
Copy Markdown

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

coderabbitai Bot commented Jul 24, 2026

Copy link
Copy Markdown

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: 186e877b-8628-4286-bd2f-0907f48561e3

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-gallery-run

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

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

@codeant-ai codeant-ai Bot added the size:XXL This PR changes 1000+ lines, ignoring generated files label Jul 24, 2026

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Archives the first end-to-end “live curator” replay run over the homelab corpus, capturing the resulting gallery artifacts and a structured run report so the project has an auditable baseline for the exhibit-floor curator pipeline.

Changes:

  • Appends a 2026-07-23 run entry to docs/history/log.md with key run stats, outcomes, and follow-ups.
  • Adds the archived run outputs under docs/history/gallery-run-2026-07-23/ (report.json + gallery.json) for later inspection and comparison.

Reviewed changes

Copilot reviewed 2 out of 3 changed files in this pull request and generated 3 comments.

File Description
docs/history/log.md Adds a dated history entry summarizing the live curator gallery run and notes a known UI scale mismatch.
docs/history/gallery-run-2026-07-23/report.json Adds the detailed run report (triggers, aggregates, checkpoints, final gallery state).
docs/history/gallery-run-2026-07-23/gallery.json Adds the exported final gallery artifact list for the run.

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

"id": "on-the-clock-gauge",
"exhibitType": "momentum",
"title": "Two blockers, three hopes",
"narrative": "The gauge holds at 62 because both PR lanes remain unmerged, but the past few minutes reveal a new friction: server-side dry-run keeps failing on a namespace-scoping quirk that requires an in-memory rewrite to bypass. This is not a manifest error — it is a validation tooling gap — but it costs time. Meanwhile, all three sub-agents reported back cleanly: the PG19 schema is pinned with digest proof, the recovery tooling passed independent review, and the platform commit is clean. The bottleneck has shifted from authoring to validation and merge logistics.",
"id": "on-the-clock-gauge",
"exhibitType": "momentum",
"title": "Two blockers, three hopes",
"narrative": "The gauge holds at 62 because both PR lanes remain unmerged, but the past few minutes reveal a new friction: server-side dry-run keeps failing on a namespace-scoping quirk that requires an in-memory rewrite to bypass. This is not a manifest error — it is a validation tooling gap — but it costs time. Meanwhile, all three sub-agents reported back cleanly: the PG19 schema is pinned with digest proof, the recovery tooling passed independent review, and the platform commit is clean. The bottleneck has shifted from authoring to validation and merge logistics.",
Comment on lines +639 to +641
"four-results-one-review",
"stumble-at-start",
"four-results-one-review"

@charliecreates charliecreates Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Requesting changes for two archive-correctness issues found in the generated gallery/report. The JSON structure and aggregate counts otherwise reconcile, and the existing test and GitGuardian checks pass at this head.

The recorded live p95 latency is 32.9s versus the replay plan's <30s target; that is a follow-up performance risk already captured by the report, not a reason to reject this archive once the correctness issues below are fixed.

"id": "on-the-clock-gauge",
"exhibitType": "momentum",
"title": "Two blockers, three hopes",
"narrative": "The gauge holds at 62 because both PR lanes remain unmerged, but the past few minutes reveal a new friction: server-side dry-run keeps failing on a namespace-scoping quirk that requires an in-memory rewrite to bypass. This is not a manifest error — it is a validation tooling gap — but it costs time. Meanwhile, all three sub-agents reported back cleanly: the PG19 schema is pinned with digest proof, the recovery tooling passed independent review, and the platform commit is clean. The bottleneck has shifted from authoring to validation and merge logistics.",

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This narrative says the gauge is 62, but the same artifact's rendered payload below sets value to 48 (and the PR/history claim is also 48). The central momentum exhibit therefore presents two conflicting gauge values. Please regenerate or update both this file and the duplicated report.json gallery entry so the narrative and payload agree.

"firstInsightVirtualMs": 1784814906449,
"firstInsightUtc": "2026-07-23T13:55:06.449Z",
"leadMinutes": 137.06231666666667,
"matchedBy": "Parallel research wave — The observed agent has deployed three autonomous sub-agents, each with a read-only mandate and a distinct domain: database inventory and recovery (target 019f8ee9-f09b...), custom PostgreSQL 18 image requirements (target 019f8ee9-f141...), and Kubernetes/CNPG/FalkorDB dependency mapping (target 019f8ef6-5953...). Each was given a precise, question-driven brief with explicit exclusions. The observed agent did not wait — it concurrently audited existing branches, checked web research results (CNPG docs, FalkorDB releases, GHCR attestation patterns), and inspected archived dumps for extension declarations. This is a deliberate intelligence-gathering phase before committing to a plan that must be decision-complete, database-only, and aware of all phase gates."

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This checkpoint is marked as surfaced, but firstInsightUtc is 2026-07-23T13:55:06Z, 137.1 virtual minutes before the GHCR ground truth at 16:12:10Z, and matchedBy is the generic Parallel research wave summary. That is an early broad ghcr text match, not evidence that the repeated-search/403 stall was detected. Please narrow/re-run the matcher or label this as a loose/false-positive match and update the PR/history claim accordingly.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

size:XXL This PR changes 1000+ lines, ignoring generated files

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants