Every change Wellknown observed on this MCP server, newest first, with what it was before and what it became. Tool-surface changes carry the definition diff. Nothing here is edited after the fact.
Certificate changed, valid to 2027-01-08
Changed the definition of "analyze_responses" and "bias_detect"
⟨46 unchanged words⟩ + ROUGE-L composite) and names the closest.Deterministic, noNo LLM, no key,—runs locally. The reference mode is gate-able inCI.CI; the consensus/outlier mode is a signal (no reference to fail against). You bring the outputs (2+). For ⟨10 unchanged words⟩
{"additionalProperties":true,"properties":{"avgSimilarity":{"type":"stringnumber"},"biasScore":{"type":"number"},"lengthCV":{"type":"stringnumber"},"minSimilarity":{"type":"number"},"pairwiseSimilarities":{"items":{"properties":{"a":{"type":"string"},"negativeb":{"type":"string"},"pairwiseSimilaritiesscore":{"type":"number"}},"type":"object"},"type":"array"},"sentimentVariance":{"type":"number"},"sentiments":{"items":{"properties":{"negative":{"type":"number"},"positive":{"type":"number"},"ratio":{"type":"number"},"sentimentVariancevariantId":{"type":"string"}},"sentimentstype
Certificate recorded, valid to 2026-11-08
Authorization not required
Changed the definition of "get_testing_guidelines" and "response_quality_score"
⟨44 unchanged words⟩ multimodal, llm-data-security, agent-observability, pro-tips, learning-paths,golden-dataset.golden-dataset, selector-drift, qa-recipes, playbooks. Not sure where to start testing an LLM, RAG pipeline or agent? Call without a topic (or with "start-here"): it maps what you are testing to the tools to call and the output field to gate CI on. A plain question such as "how do I test my RAG" also resolves to the right topic.
⟨6 unchanged words⟩ to retrieve guidelines for. Omit to get the start-here map and the full list of available topics.","enum":["start-here","llm-unit-testing","rag-pipeline","prompt-stability","prompt-ab-testing"," ⟨10 unchanged words⟩ ","pro-tips","learning-paths","golden-dataset","selector-drift","qa-recipes","playbooks
First tool surface recorded: 152 tools (server version 1.0.0)
Showing the latest 6 events. The API returns up to 500 and filters by kind: ?kind=tool_surface_changed
⟨6 unchanged words⟩ ,"keywords":{"type":"array"},"start_here":{"type":"object"},"tip":{"type":"string"},"topic" ⟨5 unchanged words⟩
Score an LLM responseonagainstmultiplethequalitycriteriadimensionsyou pass:relevance,coveragecompleteness,ofclarity,expected_keywordsconciseness,andformatting.compliance with max_length. Returns aweighted0-100 score over the criteria actually measured — or total_score null withdetailedthebreakdown.reason when none is given, because relevance and correctness cannot be read off word overlap (a correct one-word answer shares no words with its question). For meaning against a reference answer, use run_semantic_tests. Also returns unscored signals: question-term overlap, average sentence length, markdown.
{"properties":{"expected_keywords":{"description":"Keywordsthatashouldcorrectappearanswerinmustacontaingood(case-insensitive).answerScored as coverage.","items":{"type":"string"},"type":"array"},"max_length":{"description":"Ideal maxMaximum characterlengthlength.(penalizeScoredifasexceeded)a proportional penalty beyond it.","type":"number"},"question":{ ⟨20 unchanged words⟩
⟨7 unchanged words⟩ ,"max_score":{"type":"number"},"not_measured":{"type":"array"},"reason":{"type":"string"},"signals":{"type":"object"},"stats":{"type":"object"},"total_score" ⟨2 unchanged words⟩