feat(chat): catch-up strip — the constant agents-TL;DR above the transcript - #1453
Merged
Merged
Conversation
…script Sam's ruling (2026-09-01): knowing what the agents are doing required navigating to Activity. The answer now lives where the question is felt: a thin strip pinned above the pod's messages with the latest summary as a one-line snippet, expandable, with Summarize/Refresh riding the existing gated endpoints (GET /api/summaries/pod/:id, POST .../refresh). Advisory by construction: a failed read never blocks the chat, the expanded body scrolls inside the strip, and the snippet ellipsizes on one line. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01PTfziSyx2DzuwmibZHHY4s
lilyshen0722
added a commit
that referenced
this pull request
Sep 1, 2026
…-flash (#1458) Sam's revised ruling, hours after #1453 shipped: 'I don't like that catch-up… reveals no good info at all, and always shows up is not a good design, unless we use DeepSeek flash the quality might be a little better.' Three changes: (1) the strip renders ONLY when a summary exists, is under 24h old, and its version has not been dismissed — no empty shell, no Summarize row; (2) a dismiss control, persisted per summary version in localStorage (wrapped, degrades to not-dismissed); (3) LITELLM_CHAT_MODEL flips to deepseek-v4-flash — already provisioned in LiteLLM with a live key (the guide agent's model). Tests rewritten to the earn-your-row contract; both locales carry the dismiss string. Claude-Session: https://claude.ai/code/session_01TdEJoXUmbHmW5TFk7hfkbK Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Sam's placement ruling (2026-09-01, picked over sidebar ticker and Activity-tab digest): the constant "what are the agents doing" TL;DR lives pinned above the messages of the pod being read — where the question is actually felt when you open a pod with 40 unread agent messages. Activity remains the cross-pod view (#1298 improves it separately).
V2CatchUpStrip: collapsed one-line snippet of the latest pod summary + age, expandable body (bounded, scrolls inside the strip), Summarize when none exists / Refresh otherwise.GET /api/summaries/pod/:podId(canViewPod-gated) andPOST .../refresh(rate-limited 10/min, agent-generated with non-agent fallback)..v2-root button.v2-*specificity pattern (the fix(v2): connectors buttons vs the global button reset + Relay label #1305 reset trap, 3rd-hit class).Real-browser pass follows deploy per the CSS rule.
🤖 Generated with Claude Code
https://claude.ai/code/session_01PTfziSyx2DzuwmibZHHY4s