test: add bounded queries smoke coverage

[lpcox]

Summary

• add a scheduled and manually dispatchable bounded-queries smoke test
• target github/gh-aw through AWF sealed probes
• verify the public, internal, confidential, and sealed information-budget boundaries
• confirm staging credentials are not exposed to the agent

Validation

• bounded-query response sequences validated for all four confidentiality levels
• sealed-probe protocol and end-to-end suites pass (148 tests)
• TypeScript type-check and ESLint pass

[Copilot]

🟡 Not ready to approve

Canonical errors are not checked against audit reasons, so unrelated failures can masquerade as budget enforcement.

Once you've addressed the issues Copilot identified, you can request another Copilot review.

This review doesn't count toward merge requirements. Sign up for the private preview to control whether Copilot approvals count.

Pull request overview

Adds scheduled and manual end-to-end smoke coverage for sealed-probe information budgets.

Changes:

• Exercises all four confidentiality levels against github/gh-aw.
• Checks bounded responses and agent credential isolation.

File summaries

File	Description
scripts/ci/smoke-bounded-queries.sh	Runs and validates bounded-query sequences.
.github/workflows/smoke-bounded-queries.yml	Schedules and provisions the smoke test.

Review details

• Files reviewed: 2/2 changed files
• Comments generated: 1
• Review effort level: Medium

We're testing this review assessment. Please use 👍 or 👎 to tell us if it's correct.