Add AI event summary highlights on detail pages#247
Open
chrisballinger wants to merge 5 commits intomasterfrom
Open
Add AI event summary highlights on detail pages#247chrisballinger wants to merge 5 commits intomasterfrom
chrisballinger wants to merge 5 commits intomasterfrom
Conversation
Remove hidesBottomBarWhenPushed from detail screens, tracks, AI guide, recently viewed, and feature flags views for a consistent Liquid Glass tab bar experience. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Show a concise AI-generated pro-tip summarizing hosted events on camp, art, and event detail pages, plus the hosted events list view. Uses the existing workflow pipeline (retryWithCandidateFiltering + withContextWindowRetry) so summaries work even when individual events trigger guardrails or exceed the context window. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Instead of regenerating the AI summary, pass it through from the detail page. Also move the summary into the scrollable list content so it scrolls away with the rest of the rows. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Split AI summary into Swift-generated schedule tips (factual, from real occurrence data) and LLM-generated overview (no timing info allowed). This prevents the on-device model from fabricating event times. - Schedule tips: group by event name (fixes duplicates), sorted by day-of-week, strikethrough for expired (only during festival), tappable to navigate to event detail, themed highlight color - RAM cache (actor-based EventSummaryCache) keyed by host UID - Model pre-warming at AgentOrchestrator init - LLM overview: 1-2 sentences, can mention event names, no times/days - Host camp description passed to LLM for richer context - Tips show instantly, LLM overview arrives async Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Pass 1 generates the overview with tightened instructions ("only
reference provided data"). Pass 2 validates against the exact source
data (camp description + event names/descriptions) using a separate
LLM session that flags unsupported claims. Swift strips flagged
phrases. If validation fails or strips too much, the overview is
discarded entirely — no unverified text shown.
Each step handles its own retries via withContextWindowRetry /
retryWithCandidateFiltering, matching the pattern used by other
AI workflows.
Co-Authored-By: Claude Opus 4.6 (1M context) <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.
Summary
retryWithCandidateFiltering+withContextWindowRetry) for graceful handling of guardrail violations and context overflowTest plan
🤖 Generated with Claude Code