test: add unit tests for eval-outcomes, objective-default-models, and mock-recommendations#11519
test: add unit tests for eval-outcomes, objective-default-models, and mock-recommendations#11519roomote[bot] wants to merge 1 commit intofeat/eval-recommendationsfrom
Conversation
… mock-recommendations
Reviewed 3 new test files (28 tests). One suggestion:
Mention @roomote in a comment to request specific changes to this pull request or fix all unresolved issues. |
| import type { EvalOutcomeId } from "../eval-outcomes" | ||
|
|
||
| const ALL_OUTCOME_IDS: EvalOutcomeId[] = [ | ||
| "prototype_to_pr", | ||
| "paper_cuts", | ||
| "sentry_triage", | ||
| "repro_to_fix", | ||
| "review_guardrails", | ||
| "issue_to_pr", | ||
| ] |
There was a problem hiding this comment.
ALL_OUTCOME_IDS is hardcoded here rather than derived from the source of truth. TypeScript enforces that each listed value is a valid EvalOutcomeId union member, but it does not enforce completeness -- if a new outcome is added to the union type, this array won't cause a compile error and the new outcome silently won't be tested. Consider deriving the list from EVAL_OUTCOMES.map(o => o.id) (imported from ../eval-outcomes) so coverage stays in sync automatically.
| import type { EvalOutcomeId } from "../eval-outcomes" | |
| const ALL_OUTCOME_IDS: EvalOutcomeId[] = [ | |
| "prototype_to_pr", | |
| "paper_cuts", | |
| "sentry_triage", | |
| "repro_to_fix", | |
| "review_guardrails", | |
| "issue_to_pr", | |
| ] | |
| import { EVAL_OUTCOMES } from "../eval-outcomes" | |
| const ALL_OUTCOME_IDS = EVAL_OUTCOMES.map((o) => o.id) |
Fix it with Roo Code or mention @roomote and request a fix.
|
🚀 Preview deployed! Your changes have been deployed to Vercel: Preview URL: https://roo-code-website-kv8scrly0-roo-code.vercel.app This preview will be updated automatically when you push new commits to this PR. |
Adds test coverage for the new utility modules introduced in #11444:
EVAL_OUTCOMESdata integrity,isEvalOutcomeId(), andgetEvalOutcomeBySlug()pickObjectiveDefaultModelV1()across all outcome/mode combinations, verifying quality floor, speed, and cost optimization invariantsgetEngineerRoles(),getEngineerRole(),getRoleRecommendation(),getAllRecommendations(),getCloudSetupUrl(),TASKS_PER_DAY, andMODEL_TIMELINE28 tests total, all passing.
View task on Roo Code Cloud
Start a new Roo Code Cloud session on this branch