Skip to content

Retarget default RSI visible eval from rsi to engine suite #145

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
w4ffl35 merged 1 commit into from
Oct 13, 2026
Merged

Retarget default RSI visible eval from rsi to engine suite #145

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion src/rsi/__tests__/proof-demonstration-launcher.test.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -117,7 +117,7 @@ test("proof max-token option rejects invalid values and commands governed by a f
test("proof preview selects a nonempty visible test set outside withheld RSI sources", async () => {
const parsed = parseProofDemonstrationArgs(["engineering", "--seed", "preview-seed"])
const defaultConfig = (await import("../config.js")).parseRsiArgs(["--repo", process.cwd()]).config!
assert.deepEqual(defaultConfig.evalCommands, ["npm test -- --filter rsi"])
assert.deepEqual(defaultConfig.evalCommands, ["npm test -- --filter engine"])
const previewConfig = configureProofVisibleEvaluations(defaultConfig)
assert.equal(previewConfig.regressionCommand, "npm test", "retain the full regression command")
assert.deepEqual(previewConfig.evalCommands, [...PROOF_VISIBLE_EVAL_COMMANDS])
Expand Down
6 changes: 5 additions & 1 deletion src/rsi/config.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,11 @@ import type { ComputePolicy, MutationHypothesis, MutationKind, ParentSelectionPo
import { DEFAULT_DEVELOPMENT_PROOF_RULE } from "./development-proof.js"

export const DEFAULT_RSI_MODEL = "wxrq-qwen3.5-9b:latest"
export const DEFAULT_VISIBLE_EVALS = ["npm test -- --filter rsi"]
// Fleet eval snapshots always omit src/rsi/ (alwaysOmit in openshell.ts — the
// supervisor/evaluator tree is withheld from guests), so a visible eval
// targeting rsi tests matches zero files and fails every baseline. Target
// the engine suite instead (same choice as PROOF_VISIBLE_EVAL_COMMANDS).
export const DEFAULT_VISIBLE_EVALS = ["npm test -- --filter engine"]
export const DEFAULT_PROTECTED_PATHS = [
...DEFAULT_PROTECTED_FILES,
"src/rsi/",
Expand Down
Loading