Deterministic test-results health analysis: flaky/ever-failing/slowest/failure-cluster metrics from JUnit XML or CSV, with optional JSON trace correlation (correlate failing tests to concurrent activity in a companion trace file). Returns {"summary", "flaky_tests", "ever_failing", "slowest", "failure_clusters", "duration_ms"}; no LLM anywhere in this call path, write your own narrative from the numbers.

Deterministic QA process maturity assessment (TMMi/EU AI Act indicative maturity level + findings) from a free-text project/process description — no LLM anywhere in this call path; write your own narrative around the maturity level returned.

Deterministic QA effort estimate (PERT + complexity multipliers + team capacity + confidence score) — no LLM narrative; write your own from these numbers. Fields mirror the app's project-intake dialogue and are validated with the same rules; a validation failure returns {"error": "validation", "fields": {field: message}}, never a crash. Success returns the full EstimationData as JSON (baseline, multipliers, pert_activities, capacity, risk_buffer_days, final_effort_min/max, confidence_level/confidence_score).

List every document in the knowledge base, grouped by category. Returns {"categories": {category: [{"source", "title"}]}, "kb_version", "doc_count"}.

Retrieve grounding chunks from the QA knowledge base (ISTQB, OWASP, IEEE, ISO standards; testing methodologies; audit/evaluation frameworks; the EU AI Act). Returns {"chunks": [{"source", "category", "text", "score"}], "kb_version"}. category, if given, must be one of: Standard, Methodology, Article, Expert Knowledge, Audit/Evaluation — an unrecognized value returns a structured {"error": "invalid_argument", ...} rather than raising. k is clamped to [1, 20].