harness: a field under either spelling, and the runtime defaults shown as what they are - #212
Merged
Merged
Conversation
…n as what they are Issue #199: a client that mirrored the shape GET returns sent `defaultModel`; the server stored nothing and answered 200 with `"defaultModel": ""`, so every task had to carry a model, and the conformance suite waited out its timeout per task, half an hour to say nothing. The schema names the create shape in snake_case and the object in camelCase; a field the server understands under one is not an unknown field under the other. HarnessBody now takes both, and the suite, given a harness with no default model and no --model, says so and stops before the first task. The runtime limits: a customer's task ran away and drained their credits, and when they looked for the limit the settings page showed a grey placeholder. The step default gets one home beside the timeout's (DEFAULT_MAX_STEP, 400), /v1/bases serves both as runtimeDefaults, and the console shows them as the fields' values with the sentence "Empty means the default, N": the number a person sees is the number that will stop the task, from the server that enforces it. A per-task cost limit is the hosted service's to add; this is the part every instance can have. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
richard-epsilla
approved these changes
Sep 18, 2026
richard-epsilla
left a comment
richard-epsilla
left a comment
Collaborator
There was a problem hiding this comment.
Approved. Verified on the Azure test VM on top of 0.18.2: the issue's exact request stores the model, the bases document serves the defaults, the suite stops with the sentence on a harness with no default model, and the settings page shows 400 and 120 with the sentences at 1440 and 390.
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.
Fixes #199. A client that mirrored the shape GET returns sent
defaultModel; the server stored nothing and answered 200 with"defaultModel": "", so every task had to carry a model, and the conformance suite waited out its timeout per task. The schema names the create shape in snake_case and the object in camelCase; a field the server understands under one spelling is not an unknown field under the other.HarnessBodynow takes both, and the suite, given a harness with no default model and no--model, says so and stops before the first task.The runtime limits: a customer's task ran away and drained their credits, and when they looked for the limit the settings page showed a grey placeholder. The step default gets one home beside the timeout's (
DEFAULT_MAX_STEP, 400),/v1/basesserves both asruntimeDefaults, and the console shows them as the fields' values with the sentence "Empty means the default, N": the number a person sees is the number that will stop the task, from the server that enforces it. A per-task cost limit is the hosted service's to add; this is the part every instance can have.Verified on the Azure test VM on top of 0.18.2: the issue's exact request stores
codex-tools;/v1/basesanswersruntimeDefaults: {maxStep: 400, timeoutSeconds: 7200}; the suite stops with the sentence on a harness with no default model; the settings page shows 400 and 120 in the fields with the sentences under them at 1440 and 390 with no sideways scroll. Gateway 336, conformance 79, console type-check clean.🤖 Generated with Claude Code