Skip to content
Merged
Changes from all commits
Commits
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
258 changes: 258 additions & 0 deletions .github/workflows/sandbox-capabilities.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,258 @@
name: Sandbox Capabilities Matrix

on:
workflow_dispatch:
inputs:
provider:
description: 'Provider to run (leave empty for all)'
required: false
default: ''
iterations:
description: 'Iterations per provider'
required: false
default: '1'
concurrency:
description: 'Concurrent sandboxes per provider'
required: false
default: '1'
publish:
description: 'Commit and push results to the repo'
required: false
default: true
type: boolean

concurrency:
group: sandbox-capabilities
cancel-in-progress: true

permissions: {}

jobs:
bench:
name: Capabilities ${{ matrix.provider }}
permissions:
contents: read
id-token: write
runs-on: namespace-profile-default;permissions.additional_grant=vault/object:*:list;permissions.additional_grant=vault/object:*:describe
# runs-on: self-hosted
timeout-minutes: 30
env:
SHOULD_RUN: ${{ github.event_name != 'workflow_dispatch' || github.event.inputs.provider == '' || github.event.inputs.provider == matrix.provider }}
strategy:
fail-fast: false
matrix:
provider:
- archil
- arker
- beam
- blaxel
- cloud-run
- cloudflare
- codesandbox
- createos
- daytona
- declaw
- e2b
- hopx
- isorun
- lightning
- modal
- mosaic
- namespace
- northflank
- opencomputer
- runloop
- run-cloud
- sandbox0
- sail
- sprites
- superserve
- tenki
- tensorlake
- upstash
- vercel
steps:
- uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
if: env.SHOULD_RUN == 'true'
with:
persist-credentials: false
- uses: pnpm/action-setup@b906affcce14559ad1aafd4ab0e942779e9f58b1 # v4
if: env.SHOULD_RUN == 'true'
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
if: env.SHOULD_RUN == 'true'
with:
node-version: 24
cache: 'pnpm'
- uses: namespacelabs/nscloud-setup@df198f982fcecfb8264bea3f1274b56a61b6dfdc # v0.0.12
if: env.SHOULD_RUN == 'true'
- name: Install dependencies
if: env.SHOULD_RUN == 'true'
run: pnpm install --frozen-lockfile
- name: Clear stale results from checkout
if: env.SHOULD_RUN == 'true'
run: rm -rf results/
- name: Run capabilities benchmark
if: env.SHOULD_RUN == 'true'
env:
ITERATIONS: ${{ github.event.inputs.iterations || '1' }}
CONCURRENCY: ${{ github.event.inputs.concurrency || '1' }}
PROVIDER: ${{ matrix.provider }}
run: |
COMMON_KEYS='BENCHMARKS_PLATFORM_API_KEY|COMPUTESDK_ADMIN_API_KEY'
case "$PROVIDER" in
archil) PROVIDER_KEYS='ARCHIL_API_KEY|ARCHIL_REGION|ARCHIL_DISK_ID' ;;
arker) PROVIDER_KEYS='ARKER_API_KEY' ;;
beam) PROVIDER_KEYS='BEAM_TOKEN|BEAM_WORKSPACE_ID' ;;
blaxel) PROVIDER_KEYS='BL_API_KEY|BL_WORKSPACE' ;;
cloud-run) PROVIDER_KEYS='CLOUD_RUN_SANDBOX_URL|CLOUD_RUN_SANDBOX_SECRET' ;;
cloudflare) PROVIDER_KEYS='CLOUDFLARE_SANDBOX_URL|CLOUDFLARE_SANDBOX_SECRET' ;;
codesandbox) PROVIDER_KEYS='CSB_API_KEY' ;;
createos) PROVIDER_KEYS='CREATEOS_SANDBOX_API_KEY' ;;
daytona) PROVIDER_KEYS='DAYTONA_API_KEY' ;;
declaw) PROVIDER_KEYS='DECLAW_API_KEY' ;;
e2b) PROVIDER_KEYS='E2B_API_KEY' ;;
hopx) PROVIDER_KEYS='HOPX_API_KEY' ;;
isorun) PROVIDER_KEYS='ISORUN_API_KEY' ;;
lightning) PROVIDER_KEYS='LIGHTNING_API_KEY' ;;
modal) PROVIDER_KEYS='MODAL_TOKEN_ID|MODAL_TOKEN_SECRET' ;;
mosaic) PROVIDER_KEYS='MOSAIC_API_URL|MOSAIC_API_TOKEN' ;;
namespace) PROVIDER_KEYS='NSC_TOKEN' ;;
northflank) PROVIDER_KEYS='NORTHFLANK_TOKEN|NORTHFLANK_PROJECT_ID' ;;
opencomputer) PROVIDER_KEYS='OPENCOMPUTER_API_KEY|OPENCOMPUTER_API_URL' ;;
runloop) PROVIDER_KEYS='RUNLOOP_API_KEY' ;;
run-cloud) PROVIDER_KEYS='RUN_CLOUD_API_KEY' ;;
sandbox0) PROVIDER_KEYS='SANDBOX0_TOKEN' ;;
sail) PROVIDER_KEYS='SAIL_API_KEY' ;;
sprites) PROVIDER_KEYS='SPRITES_TOKEN' ;;
superserve) PROVIDER_KEYS='SUPERSERVE_API_KEY' ;;
tenki) PROVIDER_KEYS='TENKI_API_KEY' ;;
tensorlake) PROVIDER_KEYS='TENSORLAKE_API_KEY' ;;
upstash) PROVIDER_KEYS='UPSTASH_BOX_API_KEY' ;;
vercel) PROVIDER_KEYS='VERCEL_TOKEN|VERCEL_TEAM_ID|VERCEL_PROJECT_ID' ;;
*) PROVIDER_KEYS='' ;;
esac
if [ -n "$PROVIDER_KEYS" ]; then
KEYS="^(${COMMON_KEYS}|${PROVIDER_KEYS})="
else
KEYS="^(${COMMON_KEYS})="
fi
. benchmarks/scripts/load-vault-secrets.sh "$KEYS"

RUN_KEY="${GITHUB_RUN_ID}-${GITHUB_RUN_ATTEMPT}"
pnpm exec -- tsx packages/benchsdk-runner/dist/bin.js run benchmarks/sandbox/capabilities.bench.ts \
--provider "$PROVIDER" \
--iterations "$ITERATIONS" \
--concurrency "$CONCURRENCY" \
--run-key "$RUN_KEY"
- name: Upload results
if: always() && env.SHOULD_RUN == 'true'
uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
with:
name: results-capabilities-${{ matrix.provider }}
path: results/
if-no-files-found: ignore
retention-days: 7

collect:
name: Collect Capabilities Results
runs-on: namespace-profile-default
# runs-on: self-hosted
needs: bench
if: always()
permissions:
contents: write
steps:
- uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
with:
fetch-depth: 0
- uses: pnpm/action-setup@b906affcce14559ad1aafd4ab0e942779e9f58b1 # v4
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
with:
node-version: 24
cache: 'pnpm'
- uses: namespacelabs/nscloud-setup@df198f982fcecfb8264bea3f1274b56a61b6dfdc # v0.0.12
- name: Install dependencies
run: pnpm install --frozen-lockfile
- name: Download all results
uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
with:
path: artifacts/
pattern: results-capabilities-*
- name: Merge results
continue-on-error: true
run: pnpm exec -- tsx benchmarks/src/merge-results.ts --input artifacts
- name: Write GitHub Step Summary
if: always()
uses: actions/github-script@f28e40c7f34bde8b3046d885e986cb6290c5673b # v7.1.0
with:
Comment thread
kisernl marked this conversation as resolved.
script: |
const fs = require('fs');
const path = require('path');

const latestPath = path.join('results', 'sandbox-capabilities', 'latest.json');
if (!fs.existsSync(latestPath)) {
core.summary.addHeading('Sandbox Capabilities Matrix', 2);
core.summary.addRaw('No capabilities results were generated.');
await core.summary.write();
return;
}

const data = JSON.parse(fs.readFileSync(latestPath, 'utf-8'));
const results = data.results || [];
const featureNames = results.length > 0 ? Object.keys(results[0].features || {}) : [];

const summary = core.summary;
summary.addHeading('Sandbox Capabilities Matrix', 2);
summary.addRaw(`<p>Timestamp: ${data.timestamp || 'unknown'}<br>Providers: ${results.length}</p>`);

const header = ['Provider', ...featureNames];
const rows = results.map(({ provider, features }) => {
return [
provider,
...featureNames.map((name) => (features && features[name] && features[name].passed ? '✅' : '❌')),
];
});
summary.addTable([header, ...rows]);

const failures = results
.map(({ provider, features }) => ({
provider,
failed: Object.entries(features || {})
.filter(([, value]) => !value.passed)
.map(([name, value]) => ({ name, error: value.error || 'failed' })),
}))
.filter((entry) => entry.failed.length > 0);

if (failures.length > 0) {
summary.addHeading('Failure details', 3);
for (const { provider, failed } of failures) {
const content = failed
.map((failure) => `- **${failure.name}**: ${failure.error}`)
.join('\n');
summary.addDetails(`${provider} (${failed.length} failures)`, content);
}
}

await summary.write();
- name: Commit and push
if: github.event.inputs.publish != 'false'
run: |
git config user.name "github-actions[bot]"
git config user.email "github-actions[bot]@users.noreply.github.com"
mkdir -p results/sandbox-capabilities
git add results/sandbox-capabilities/
git diff --cached --quiet && echo "No changes to commit" && exit 0
Comment thread
kisernl marked this conversation as resolved.
git commit -m "chore: update sandbox capabilities results [skip ci]"

branch="${GITHUB_REF#refs/heads/}"
for attempt in 1 2 3 4 5; do
git fetch origin "${branch}"
git rebase --autostash "origin/${branch}" || { git rebase --abort; exit 1; }
if git push origin "HEAD:${branch}"; then
echo "Pushed on attempt ${attempt}"
exit 0
fi
echo "Push rejected (attempt ${attempt}); will rebase and retry"
done
echo "Failed to push after multiple attempts" >&2
exit 1
Loading