vikingowl 7fbb5454ee feat(router): normalize effort/thinking abstraction across providers ...

Add EffortLevel (auto/low/medium/high) as a provider-agnostic reasoning
control, replacing the Capabilities.Thinking bool. Each provider maps
the level to its native parameter: Anthropic budget tokens (1K/8K/16K),
OpenAI reasoning_effort (low/medium/high), Google thinking budget
(1K/8K/16K). Task classification auto-infers effort from TaskType and
complexity; filterFeasible excludes arms that lack the required level.

2026-05-07 14:08:50 +02:00

..

app.go

feat(router): normalize effort/thinking abstraction across providers

2026-05-07 14:08:50 +02:00

completions_test.go

feat(tui): Tier 1-2 UX improvements — completions, usage, provider status

2026-04-12 04:19:55 +02:00

completions.go

feat(tui): Tier 3-4 UX improvements — split, routing, session naming, context bar

2026-04-12 05:13:16 +02:00

events.go

feat(tui): Tier 3-4 UX improvements — split, routing, session naming, context bar

2026-04-12 05:13:16 +02:00

init.go

feat: local model reliability — SDK retries, capability probing, init skill, context compaction

2026-04-13 02:01:01 +02:00

rendering.go

feat: local model reliability — SDK retries, capability probing, init skill, context compaction

2026-04-13 02:01:01 +02:00

statusbar_test.go

feat(tui): Tier 3-4 UX improvements — split, routing, session naming, context bar

2026-04-12 05:13:16 +02:00

theme.go

feat: Ollama/gemma4 compat — /init flow, stream filter, safety fixes

2026-04-05 19:24:51 +02:00