Review the complete bound workspace; poll with continue_review until terminal.

Wait briefly for a long review; repeat until state is completed or failed.

Start Guardian high only after a completed first review of the same source snapshot.

Scan a local workspace for secrets, Korean PII, composite PII, config risks, and flow risks.

Scan an inline text snippet without writing it to disk.

Scan only a git diff, without echoing raw diff content in the MCP response.

Discover and scan project MCP configs plus agent instruction and skill components.

Run safe read-only HTTP probes against localhost or explicitly authorized external targets.

Build an EXPERIMENTAL heuristic data-flow risk map for source and sink patterns.

Observe MCP runtime/proxy JSONL events for PII, hidden instructions, and agentic/external flow.

Evaluate MCP runtime block/redact policy and return a raw-free report-only interceptor result.

Observe one MCP runtime event and return stream-ready block/redact policy decisions.

Evaluate a local fixture corpus and return FP/FN precision/recall scoreboard.

Run the local fail-closed Semgrep deep profile; this is intrafile evidence, not an authorization proof.

Run the pinned nine-language pack twice; this does not claim field accuracy.

Audit Python/npm/Go locks; explicit authorization is required because engines may consult advisory databases.

Run the repeated synthetic JIT/JEA and database AST/RBAC/isolation control pack.

Run the repeated Streamable HTTP authorization, lifecycle, SSE, and audit matrix.

Run the strict data shipment gate over Guardian, validation, Korean corpus, and MCP interceptor evidence.

Create the empty 12-20 app owned/partner roster used before holdout preregistration.

Validate the app roster and emit a raw-free fail-closed campaign status.

Create a 20/20/10 field benchmark manifest CSV template.

Aggregate field benchmark reports by cohort; probes require explicit MCP probe opt-in.

Create a korean_senior Guardian manifest for four-domain pre-release reviews.

Run Guardian; korean_senior rows require four review domains and fail closed when fail_on is set.

Explain a K-Guard rule in plain language with impact, fix steps, and verification.

Return a dry-run fix recipe and regression-test suggestions for a rule.

Run the quick workspace gate. Use guardian_audit(fail_on=...) for the canonical release gate.