mirror of
https://github.com/garrytan/gstack.git
synced 2026-10-31 22:14:25 +01:00
* feat(evals): Opus 5.5 is the paid gate's model under test; judge stays Fable (GSTA-49 TM1) Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com> * v1.91.140.0 chore: stamp TM1 (Opus 5.5 model under test) and CHANGELOG Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com> * fix(evals): repair the five Opus gate reds from run 39161502626 - coverage audit: owned cd; followed by one && chain, wc -l on in-cwd files - shared-libs: Claude Code 2.1.292 git --version probe, exact argv only - ship hook: accept 'integrate the hook yourself' / 'combine ... by hand' - qa-only: deadline starts only after the charter Write completes - plan floor: tool-only parent session identifies the owned Create grant; floor runner fails fast with PERMISSION_DIALOG_UNANSWERED Each repair replays its captured run as a free regression with negative controls. HARNESS_VERSION stays 6 (re-recorded from main's 5). Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com> * fix(evals): ship hook manual-integration sentence rule; qa-only inspection uses Read/Glob; tool-only permission tests join their owner file Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com> * fix(evals): tpa-absent-linux accepts instructions the user follows themselves as the manual option Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com> * fix(evals): owned plan Edit menu binds to exactly one pending Edit, including unpublished ones Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com> * fix(evals): shared-libs read evidence counts a literal path loop whose output names each file Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com> * chore: restamp to v1.91.142.0 (queue slot moved) Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com> * fix(evals): investigate boundary fixture identifies the owner question by its continue/abort options Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com> * fix(evals): plan-design-with-ui focus check accepts the skill's design-completeness rating in the question body Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com> * fix(evals): phrasing sweep across the 108 gate rule cases; replay captured Opus paraphrases with negative controls Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com> * fix(evals): command-detector sweep — exact-count anchor, label slot, 2.1.292 worktree probe - coverage audit (gate 39207508651 plan-eng-coverage-audit): output of exactly known length (wc -l over literal in-cwd files) anchors the next owned read at that line; a label's slot counts from the end and a later complete copy voids the credit (review/specialists/testing.md prints '---') - shared-libs-unsupported-git (39161502626): Claude Code 2.1.292's askPass-first 'worktree list --porcelain -z' cwd probe is Claude's own; the read-only Git check is shared (repoGitRequests, isReadOnlyRepoGitRequest) and the free regression replays every Git request of that capture Each fix replays its captured command with negative controls. No threshold, trial count, case kind or judge changed. HARNESS_VERSION stays 6. Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com> --------- Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
59 lines
2.5 KiB
TypeScript
59 lines
2.5 KiB
TypeScript
/**
|
|
* Host-neutral eval/harness model resolution (fork port wave 2, G cluster).
|
|
*
|
|
* Model IDs were hardcoded at six call sites across the eval helpers and the
|
|
* distill bin, so an environment that pins a different model (CI cost
|
|
* control, a model migration, an air-gapped proxy alias) had to patch source.
|
|
* One resolution point, env-overridable:
|
|
*
|
|
* GSTACK_EVAL_MODEL_<KIND> (e.g. GSTACK_EVAL_MODEL_WARMUP) — per-kind
|
|
* GSTACK_EVAL_MODEL — global
|
|
* explicit argument — caller wins
|
|
* per-kind default — last resort
|
|
*
|
|
* Kinds and their defaults:
|
|
* capture — the model under test in every Claude session, PTY and Agent SDK
|
|
* eval run: CLAUDE_CAPTURE_EVAL_MODEL
|
|
* warmup — PTY warm-up ping (cheapest thing that answers): haiku
|
|
* distill — free-text distillation (cheap, structured): haiku (pinned)
|
|
* judge — LLM-judge rubric calls: CLAUDE_JUDGE_EVAL_MODEL
|
|
*/
|
|
|
|
/** The model under test: Anthropic's best code model (Garry, 2026-10-18). */
|
|
export const CLAUDE_CAPTURE_EVAL_MODEL = "claude-opus-5-5";
|
|
/**
|
|
* The judge is the ruler, not the model under test: its calibration fixtures
|
|
* (test/fixtures/judge-calibration) are keyed to it, so it moves only with a
|
|
* recalibration, never with the model under test.
|
|
*/
|
|
export const CLAUDE_JUDGE_EVAL_MODEL = "claude-fable-5-1";
|
|
export const DEFAULT_JUDGE_MAX_TOKENS = 8192;
|
|
|
|
// `as const satisfies` keeps EvalModelKind the literal union
|
|
// 'capture' | 'warmup' | 'distill' — a `Record<string, string>` annotation
|
|
// would widen it to string and let any typo through the type gate.
|
|
const DEFAULTS = {
|
|
// Tests needing a cheaper or historical model pass it explicitly or set
|
|
// GSTACK_EVAL_MODEL_CAPTURE / GSTACK_EVAL_MODEL_JUDGE / GSTACK_EVAL_MODEL.
|
|
capture: CLAUDE_CAPTURE_EVAL_MODEL,
|
|
warmup: "claude-haiku-4-5",
|
|
distill: "claude-haiku-4-5-20251001",
|
|
judge: CLAUDE_JUDGE_EVAL_MODEL,
|
|
} as const satisfies Record<string, string>;
|
|
|
|
export type EvalModelKind = keyof typeof DEFAULTS;
|
|
|
|
export function resolveEvalModel(
|
|
kind: EvalModelKind,
|
|
explicit?: string | null,
|
|
env: NodeJS.ProcessEnv = process.env,
|
|
): string {
|
|
if (explicit) return explicit;
|
|
const perKind = env[`GSTACK_EVAL_MODEL_${kind.toUpperCase()}`];
|
|
if (perKind) return perKind;
|
|
if (env.GSTACK_EVAL_MODEL) return env.GSTACK_EVAL_MODEL;
|
|
const fallback = DEFAULTS[kind];
|
|
if (!fallback) throw new Error(`resolveEvalModel: unknown kind "${kind}"`);
|
|
return fallback;
|
|
}
|