Skip LLM for description when the page already provides one#13

Open

biafra23 wants to merge 20 commits into

mainfrom

claude/description-short-circuit

Commits on Apr 29, 2026

Replace AI router debug Toast with bottom auto-hiding status strip
biafra23
committed
Wire HF token end-to-end and fix gated-model 401s
biafra23
committed
Address PR #12 review: monotonic timing, robust Completed emit, DROP_OLDEST
biafra23
committed
Disable vision/audio backends in LiteRT-LM EngineConfig
biafra23
committed
Reject LEAP degenerate description / title output
biafra23
committed
Fire AI tags even when description generation fails
biafra23
committed
Re-trigger share-intent AI when aiCoreEnabled flips on after a delay
biafra23
committed
Append debug provenance tag to AI-generated tag lists
biafra23
committed
Debug provenance tag now shows generation duration, short SDK name
biafra23
committed
Drop dbg: prefix from debug provenance tag
biafra23
committed
Include model name in debug provenance tag
biafra23
committed
Show LiteRT-LM (not MEDIAPIPE) as runtime label in comparison rows
biafra23
committed
Handle system back from Settings / AddEdit / Comparison
biafra23
committed
Self-heal LiteRT-LM 'no OpenCL on this device' failure
biafra23
committed
Don't reload+retry LiteRT-LM inline — OOMed Pixel 7a at 5.96 GB
biafra23
committed
Prefer CPU over GPU in LiteRT-LM backend strategy
biafra23
committed
LiteRT debug tag now shows backend label; restore NPU→GPU→CPU strategy
biafra23
committed
Mode-aware dedup so AI doesn't fire twice when toggle flips on
biafra23
committed
Re-deprioritise GPU below CPU; OpenCL bug confirmed on Tensor G5 too
biafra23
committed
Skip LLM for description when the page already provides one
biafra23
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Skip LLM for description when the page already provides one#13

Skip LLM for description when the page already provides one#13
biafra23 wants to merge 20 commits into
mainfrom
claude/description-short-circuit

Commits on Apr 29, 2026

Replace AI router debug Toast with bottom auto-hiding status strip

Wire HF token end-to-end and fix gated-model 401s

Address PR #12 review: monotonic timing, robust Completed emit, DROP_OLDEST

Disable vision/audio backends in LiteRT-LM EngineConfig

Reject LEAP degenerate description / title output

Fire AI tags even when description generation fails

Re-trigger share-intent AI when aiCoreEnabled flips on after a delay

Append debug provenance tag to AI-generated tag lists

Debug provenance tag now shows generation duration, short SDK name

Drop dbg: prefix from debug provenance tag

Include model name in debug provenance tag

Show LiteRT-LM (not MEDIAPIPE) as runtime label in comparison rows

Handle system back from Settings / AddEdit / Comparison

Self-heal LiteRT-LM 'no OpenCL on this device' failure

Don't reload+retry LiteRT-LM inline — OOMed Pixel 7a at 5.96 GB

Prefer CPU over GPU in LiteRT-LM backend strategy

LiteRT debug tag now shows backend label; restore NPU→GPU→CPU strategy

Mode-aware dedup so AI doesn't fire twice when toggle flips on

Re-deprioritise GPU below CPU; OpenCL bug confirmed on Tensor G5 too

Skip LLM for description when the page already provides one