Repository navigation
Support effort on codex, pi and opencode, and in settings - #299
Merged
Merged
Conversation
adamw
changed the title
A typed model@level in the wizard now replaces an existing effort. BackendTag.harnessName replaces AgentSpec.harnessNameFor, AgentSpec.ValueForm holds the settings value form, parse is split into named steps, and docs are corrected (cheap keeps effort when the model doesn't change).
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Follow-up to #298:
withEffortnow also works on codex, pi and opencode, and effort can be set in settings andorca config.In code
EffortgetsOffandMinimal.withEffortis now a method onAgent. It needs anEffortSupportfor the harness, so on gemini it still does not compile. (Separate per-harness extensions would clash after erasure once exported fromorca.)EffortSupportlists, per harness, the levels it accepts and how each is sent. A level the harness does not accept throws when the agent is built.LowtoMax--effort-c model_reasoning_effort=...Max--thinkingvarianton the messageIn settings and
orca configAgent values are now
harness[:model][@effort], e.g.codingAgent = claude:sonnet@high,orca config --coding-agent codex@low.@are read as a level (off,minimal,low,medium,high,xhigh,max, lowercase), so a model id like Vertex'sclaude-sonnet-4@20250514is kept as is.@hgih), a trailing@, a level the harness does not accept, or any level on gemini is a settings error.agents:line,orca configand the shell summary show the level. The wizard has no effort step; it keeps an existing level when the harness stays the same, and asks again if a typed model ends in a level the harness rejects.Notes
--thinkingwins over amodel:levelpattern, and pi lowers a level the model lacks to the nearest one it has (checked in pi's source).