Check whether an AI answer is grounded in its context — deterministic, no LLM judge.
The io.github.nickjlamb/opengate-mcp MCP server checks whether an AI answer is grounded in its context using deterministic, gold-anchored verification. Its approach is described as evidence-grounded AI with no LLM judge, focused on evaluation and verification workflows.
Supports AI verification and hallucination detection
🚀 Use Cases
Benchmarking and llm-evaluation for grounding
RAG evaluation and rag-evaluation workflows
Regression-testing and evaluation tracking
Trustworthy-ai checks for response grounding
⚡ Developer Benefits
Model-context grounded verification for deterministic results
Evals for hallucination-detection and ai-verification
Tooling for Node.js (nodejs) and Python (python) ecosystems
⚠️ Limitations
The provided documentation excerpt emphasizes “no LLM judge” and grounded verification, but does not specify other supported evaluation methods or coverage beyond context grounding.
Evidence over plausibility. OpenGATE verifies AI systems that must justify every answer from source material — RAG pipelines, document-QA tools, legal and scientific assistants. It answers one question above all: can the system prove its answer from the evidence it was given?
The check is deterministic — no LLM-as-judge, no grader model, no six-point verdict scale. Required facts must be present, every number must trace back to the source, and when the context can't answer, the system must abstain rather than fabricate. Because it's pure logic, it's reproducible, free, and fast enough to run on every answer or gate on every commit.
As AI moves into high-stakes domains, evaluation is becoming as fundamental as automated testing is in traditional software. OpenGATE turns grounding failures into numbers you can track, and gates every prompt, model, or workflow change against a baseline — so reliability can't quietly regress.
Quick start (60 seconds)
No API key needed — the offline suite runs deterministic scorers against the bundled gold set:
bash
npx @pharmatools/opengate # run the offline evaluation suite
npx @pharmatools/opengate init # scaffold gold cases + HTTP config + a GitHub Action
Scorers never talk to a system directly — they reach it through a small adapter, so the methodology travels and only the gold set changes. In the development loop it sits where CI sits: change a prompt, model, or pipeline; the regression gate diffs the new scorecard against the baseline — improved or held deploys, regressed fails the build.
Why not DeepEval?
Use both — evals measure, OpenGATE verifies. General-purpose frameworks like DeepEval and OpenAI Evals evaluate AI systems broadly, usually with an LLM judging the output. OpenGATE verifies the narrower, harder promise: that every answer is grounded in evidence:
Provenance is first-class — does the cited passage actually exist, verbatim, in the source?
No LLM judge — scores are deterministic checks against hand-labelled gold, so they're reproducible and free to run in CI; your judgment lives in the gold set, not a grader model's.
Regression detection is first-class — every run is diffed against a per-adapter baseline; a drop fails the build.
Pair a general framework for broad quality metrics with OpenGATE to gate the grounding.
Core concepts
Gold cases — hand-labelled benchmark cases (datasets/cases/): source text, the claims that should be extracted, the sentences that should not be, and reference snippets with known-correct verdicts. Copy _template.json to add one; format in datasets/SCHEMA.md, labelling rules in datasets/LABELING-GUIDE.md.
Scorers — one module per metric family (src/scorers/):
Scorer
Mode
Measures
citation-detection
offline
per-claim citation set exact-match & Jaccard; supported-style accuracy
claim-extraction
online
precision / recall / F1 vs gold; non-claim leakage; fidelity (claim is verbatim from source)
verdict-accuracy
online
exact & adjacency accuracy on a six-point scale; passage hallucination rate; consistency; latency & token cost
redaction
online
recall on gold identifiers with leaks as named failures; over-redaction; known-gap tracking
fidelity of retrieved records vs the authority: anchor fields + structural invariants
grounding
online
generic RAG: answer-anchor recall, fabrication vs context, and abstention. The turnkey path
Offline scorers run with no API key — fast enough for every commit. Online scorers exercise a live system through an adapter.
Scorecards — every run writes results/<timestamp>.json stamped with the git SHA, so any result is reproducible and auditable. Per-model runs carry a run_model label, turning the results directory into a measured comparison (accuracy × hallucination × latency × cost).
Regression gate — --baseline saves a reference; later runs print per-metric deltas (▲/▼ in percentage points) and --ci fails the build on any drop. Baselines are per-adapter, so one system's scorecard can't clobber another's.
HTML report — add --report (or opengate report) for a self-contained dashboard: pass/fail per scorer, deltas vs baseline, every named failure. One file, no server, no dependencies.
Evaluating your own system
An adapter is one file: two base exports — onlineAvailable(), onlineConfigHint() — plus at least one complete capability (e.g. grounding → answer()). Scorers check adapter.capabilities and skip cleanly across the boundary; adapters are validated at load with messages naming every missing export.
bash
OPENGATE_ADAPTER=./adapters/my-system.mjs npm run eval:online
For REST-backed systems there's a no-code path: the bundled generic HTTP adapter reads endpoint paths and headers from opengate.http.json (${ENV} interpolation, built-in latency/token capture). Full contract and a minimal skeleton: ADAPTERS.md.
Examples
Evaluating a NIM-powered RAG agent — builds a RAG agent on an NVIDIA NIM model and gates its answers' grounding with OpenGATE, deterministically and with no LLM judge. Includes a runnable Python notebook (opengate-grounding) and a Node adapter for the CI gate.
Proven in production
Four PharmaTools products run on OpenGATE in CI — four different capability shapes, one evaluation standard. Run against RefCheckr's gold set, OpenGATE:
surfaced a silent parse-failure mode affecting ~50% of multi-claim verdicts, eliminated with enforced structured output (→ 0);
halved passage hallucination (5.8% → 2.4%) by driving a measured production model change — a decision made on numbers, not reputation;
holds claim extraction at 0.91 F1 with 0.93 recall on the committed baseline (run-to-run spread 0.86–0.94 — the splitter is an LLM), and is currently failing its own gate: 2 known non-claims leak into extraction on most runs, an open splitter issue the gate reports rather than rounds away.
Redacta — redaction capability (proof the methodology isn't QA-shaped)
Redacta wraps the @pharmatools/redacta engine, scored against synthetic UK clinical notes with gold-labelled identifiers. On its first run the eval found two real engine bugs (relation phrases swallowing nested names; apostrophe surnames dropped) — both fixed and confirmed (knownGap_closed: 2), then promoted to gold. Current scorecard: 100% recall on 25 gold identifiers, 0 leaks, no open gaps.
Patiently AI — simplify capability (faithfulness of paraphrase)
Patiently AI exercises faithfulness scoring for text that is paraphrase by design. The eval caught the simplifier dropping safety-critical specifics — an antibiotic dose vanished from a discharge summary (anchor recall 86%). A preservation rule took the next run to 100% anchor recall, 0 dropped facts, 0 fabricated numbers — a per-run measurement, not a guarantee: one frozen exp-2 capture made six days after the fix still contains a correct-but-unsourced reference range, flagged by design (RESULTS.md §6). That tail is why the eval now gates Patiently's backend in CI and re-scores the live service weekly, with fabrications failing on first occurrence.
PubCrawl — retrieval capability (the layer everything else grounds on)
PubCrawl has no model — it exercises retrieval fidelity against hand-verified anchors and structural invariants, catching parser regressions (collapsed author arrays, [object Object] leakage) that would poison every downstream citation. That OpenGATE scores a non-AI system at all is the point: evidence-grounded AI is only as trustworthy as the retrieval beneath it.
Contributions are welcome — especially gold cases (new domains, citation styles), adapters (connect your system), and scorers (new metric families). See CONTRIBUTING.md — including its support & governance expectations and AI-assisted development disclosure; open an issue to discuss large changes. Interfaces may still shift pre-1.0, and semver will signal breaking changes.
Citing OpenGATE
If you use OpenGATE in research, please cite it — citation metadata is in CITATION.cff (GitHub's "Cite this repository" button uses it).
License
MIT — because evaluation frameworks shouldn't be black boxes. If an evaluation influences deployment decisions, engineers should be able to inspect every scorer, metric, and benchmark.
