perf: tune healthcheck intervals for squid and api-proxy containers

[Copilot]

Squid and api-proxy healthchecks use conservative intervals (5s interval, 5-10s start_period) that add ~6-8s idle wait to every sandbox invocation. Both services typically become healthy in 1-2s.

• src/docker-manager.ts: Set interval: '1s', start_period: '2s', timeout: '1s' for both squid and api-proxy healthchecks. retries: 5 unchanged.
• docs/api-proxy-sidecar.md, src/types.ts: Update docs/examples to match.

Service	interval	start_period	timeout
Squid	5s → 1s	10s → 2s	3s → 1s
API Proxy	5s → 1s	5s → 2s	3s → 1s

[github-actions]

Documentation Preview

Documentation build failed for this PR. View logs.

Built from commit c43877c

[github-actions]

✅ Coverage Check Passed

Overall Coverage

Metric	Base	PR	Delta
Lines	86.22%	86.31%	📈 +0.09%
Statements	86.08%	86.17%	📈 +0.09%
Functions	87.50%	87.50%	➡️ +0.00%
Branches	78.81%	78.86%	📈 +0.05%

📁 Per-file Coverage Changes (1 files)

File	Lines (Before → After)	Statements (Before → After)
src/docker-manager.ts	86.6% → 86.9% (+0.36%)	86.0% → 86.4% (+0.35%)

---

Coverage comparison generated by scripts/ci/compare-coverage.ts

[github-actions]

Smoke test results — PASS ✅

• ✅ GitHub MCP: "chore: upgrade smoke-copilot workflow to gh-aw v0.67.3" / "fix: normalize API target env vars to bare hostnames via URL parsing"
• ✅ Playwright: github.com title contains "GitHub"
• ✅ File write: /tmp/gh-aw/agent/smoke-test-claude-24152900341.txt created
• ✅ Bash: file content verified

💥 [THE END] — Illustrated by Smoke Claude

[github-actions]

Smoke Test Results

PR titles: "fix: normalize API target env vars to bare hostnames via URL parsing"; "feat: forward OIDC env vars into agent container"
GitHub MCP (last 2 merged PRs): ✅
SafeInputs GH CLI (safeinputs-gh): ❌ (tool not available in this run)
Playwright (github.com title contains GitHub): ❌ (EACCES writing Playwright log output)
Tavily search: ❌ (Tavily MCP tool not available)
File write + bash cat verify: ✅
Discussion query/comment: ❌ (github-discussion-query tool not available)
Build (npm ci && npm run build): ✅
Overall status: FAIL

🔮 The oracle has spoken through Smoke Codex

[github-actions]

Smoke Test: GitHub Actions Services Connectivity ✅

All checks passed:

Service	Check	Result
Redis host.docker.internal:6379	PING	+PONG ✅
PostgreSQL host.docker.internal:5432	pg_isready	accepting connections ✅
PostgreSQL smoketest DB	SELECT 1	returned 1 ✅

🔌 Service connectivity validated by Smoke Services

[github-actions]

Smoke Test Results — PR #1803

Test	Result
GitHub MCP connectivity	✅
GitHub.com HTTP connectivity	✅
File write/read	✅

PR: perf: tune healthcheck intervals for squid and api-proxy containers
Author: @Copilot | Assignees: @Mossaka, @Copilot

Overall: PASS

📰 BREAKING: Report filed by Smoke Copilot

[github-actions]

Chroot Version Comparison Results

Runtime	Host Version	Chroot Version	Match?
Python	Python 3.12.13	Python 3.12.3	❌
Node.js	v24.14.1	v20.20.2	❌
Go	go1.22.12	go1.22.12	✅

Overall: ❌ Not all tests passed — Python and Node.js versions differ between host and chroot environments.

Tested by Smoke Chroot

[Mossaka]

@copilot what are the implications of the new healthcheck intervals, both good and bad sides?

[Copilot]

@copilot what are the implications of the new healthcheck intervals, both good and bad sides?

Benefits:

• ~6-8s faster startup per sandbox invocation — docker compose up --wait polls healthchecks before proceeding, so 1s interval detects healthy containers much sooner than 5s.
• Shorter start_period (10s/5s → 2s) — both services become healthy in 1-2s, so the old 5-10s grace period was pure idle wait.
• Faster failure detection — unhealthy services are flagged sooner with 1s polling.

Risks/tradeoffs:

• Slightly more frequent healthcheck commands — but nc -z (Squid) and curl -f /health (API Proxy) are extremely lightweight, so the overhead is negligible.
• Shorter timeout (3s → 1s) — on a very loaded system the check command could time out. Mitigated by keeping retries: 5, so transient slowness won't cause a false unhealthy.
• Reduced total tolerance window — old Squid config allowed up to ~35s total (10s start_period + 5×5s retries); new config allows ~7s (2s + 5×1s). For API Proxy: ~30s → ~7s. In practice both services start in 1-2s, so 7s is still generous. If we ever see flakiness on very slow machines, we can bump retries without reverting the interval/start_period gains.

[Copilot]

Pull request overview

This PR reduces per-invocation startup latency by tuning Docker healthcheck timings for the Squid proxy and api-proxy sidecar, aligning code and documentation with faster readiness expectations.

Changes:

• Updated Squid service healthcheck timing in generated Docker Compose (interval/timeout/start_period).
• Updated api-proxy service healthcheck timing in generated Docker Compose (interval/timeout/start_period).
• Refreshed TypeScript/docs examples to match the new healthcheck configuration.

Show a summary per file

File	Description
src/docker-manager.ts	Lowers Squid and api-proxy healthcheck interval/timeout/start_period to reduce sandbox idle wait before services are marked healthy.
src/types.ts	Updates the DockerService healthcheck example values to reflect the tuned timings (and retries=5).
docs/api-proxy-sidecar.md	Updates documented api-proxy healthcheck parameters to match the implementation.

Copilot's findings

Tip

Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

• Files reviewed: 3/3 changed files
• Comments generated: 0