Skip to content

feat: warm-boot snapshot for the Vercel Sandbox eval runner #201

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

Draft
Rodriguespn wants to merge 3 commits into
base: mattrossman/ai-912-spike-vercel-sandbox-runner-for-evals
Choose a base branch
from
Draft

feat: warm-boot snapshot for the Vercel Sandbox eval runner #201

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
25d3652
feat: opt-in Vercel Sandbox warm-boot snapshot for the eval runner
Rodriguespn Aug 14, 2026
79529c9
ci: reuse a prebuilt warm-boot snapshot via vars.EVAL_SNAPSHOT_ID
Rodriguespn Aug 14, 2026
57c3ac0
chore: refresh eval results
github-actions[bot] Aug 14, 2026
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
18 changes: 13 additions & 5 deletions .github/workflows/eval-refresh.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -302,16 +302,24 @@ jobs:
env:
EVAL_PAIRS: ${{ needs.prepare.outputs.pairs }}
EVAL_REVISION: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
# Optional warm-boot snapshot to reuse across runs (built out-of-band).
# Unset -> cold boots, so this can never regress the default path.
EVAL_SNAPSHOT_ID: ${{ vars.EVAL_SNAPSHOT_ID }}
shell: bash
run: |
set -euo pipefail

pnpm --filter @supabase-evals/framework eval:vercel -- \
--pairs-json "$EVAL_PAIRS" \
--revision "$EVAL_REVISION" \
--runs "${{ needs.prepare.outputs.runs }}" \
--timeout-sec "${{ needs.prepare.outputs.timeout_sec }}" \
args=(
--pairs-json "$EVAL_PAIRS"
--revision "$EVAL_REVISION"
--runs "${{ needs.prepare.outputs.runs }}"
--timeout-sec "${{ needs.prepare.outputs.timeout_sec }}"
--concurrency "${{ needs.prepare.outputs.sandbox_concurrency }}"
)
if [ -n "${EVAL_SNAPSHOT_ID:-}" ]; then
args+=(--snapshot-id "$EVAL_SNAPSHOT_ID")
fi
pnpm --filter @supabase-evals/framework eval:vercel -- "${args[@]}"

- name: Upload raw results
if: always()
Expand Down
8 changes: 2 additions & 6 deletions apps/framework/scripts/run-vercel-evals.test.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,11 +1,7 @@
import { APIError } from '@vercel/sandbox';
import { describe, expect, it } from 'vitest';
import {
isRetryableSandboxCreateError,
parsePairs,
runBounded,
tagValue,
} from './run-vercel-evals.js';
import { parsePairs, runBounded } from './run-vercel-evals.js';
import { isRetryableSandboxCreateError, tagValue } from './vercel-sandbox.js';

describe('Vercel eval controller', () => {
it('bounds concurrent work and lets independent failures settle', async () => {
Expand Down
Loading