Repository navigation
feat: select reasoning effort from request complexity - #24
Closed
BaconDroid wants to merge 4 commits into
Closed
BaconDroid wants to merge 4 commits into
BaconDroid wants to merge 4 commits into
Conversation
Repeated set_reasoning_effort calls in one turn could reset to an intermediate level, or cancel the reset entirely when the same level was set twice, because the baseline was re-read after the override. Keep the first reset target, read session history once per call with a bounded query, cache provider variants, merge guidance into an existing system prompt, and replace Zod config parsing with a small validator.
Fall back to the cached current level when the recent history window carries no variant, and read model identity and variants from assistant messages too. Reuse the SDK provider type for the cached provider list.
Add an opt-in autoSelect mode. The chat.message hook classifies each user request and writes the chosen variant onto the message, which OpenCode persists and uses for the turn. The level is carried across turns until the agent changes it or the model changes.
Respect a variant supplied with the request and treat image-only prompts as no signal. Document that autoSelect classifies the first request and carries the level instead of reclassifying every turn.
BaconDroid
force-pushed
the
feat/auto-effort
branch
from
October 19, 2026 18:11
8f6b9a1 to
6334f96
Compare
Author
|
Closing: the word-list complexity classifier this PR introduces is too heuristic. The fork is reverting to the simpler reset fix in #23. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Closes #20.
Adds an opt-in
autoSelectmode that classifies the first user request in a session and carries the chosen reasoning effort from the current model's variants.Stacked on #23. This branch is based on
fix/reliable-temporary-resets, so the diff includes #23. Review/merge #23 first; I will rebase this branch once it lands.How it works
chat.messagereceives the incomingUserMessage. OpenCode persists that object after the hook (trigger("chat.message", …, { message, parts })followed byupdateMessage(message)) andSessionPrompt.runcopiesX.model.variantonto the assistant message, so writing the chosen variant onto the message applies it to the turn without a synthetic prompt.src/complexity.tsscores the prompt for debugging, architecture, error, test, scope, and length signals, then mapssimple/normal/complex/deepto an available variant.none,off, anddisabledare skipped; single-variant models keep their only level.set_reasoning_effort, and a model switch reclassifies.In this mode the plugin manages turns that do not carry an explicit variant; explicit variants are honored.
Tests
pnpm verifypasses (21 tests, format, lint, typecheck, build).The heuristic is intentionally dependency-free and deterministic; it can be tuned without touching the plugin flow.