Build a Tiny Citation Gate Before Trusting RAG Answers

A developer built a deterministic citation gate to verify whether cited text contains enough evidence for a claim in RAG systems. The gate checks for missing sources and required terms without semantic reasoning, serving as a fast first test before more complex verification. The approach prioritizes inspectability and precision-recall tradeoffs over full semantic understanding.

A RAG answer can cite a real document and still make an unsupported claim. Retrieval answers “which text was nearby?” Citation verification asks a different question: “does the cited text contain enough evidence for this sentence?” Start with a deterministic gate before adding another model. python from dataclasses import dataclass @dataclass class Claim: text: str citation ids: list str required terms: set str def verify claim: Claim, sources: dict str, str - dict: missing = cid for cid in claim.citation ids if cid not in sources evidence = " ".join sources.get cid, "" for cid in claim.citation ids .lower absent = sorted term for term in claim.required terms if term.lower not in evidence return { "claim": claim.text, "valid source ids": not missing, "missing sources": missing, "missing terms": absent, "supported": not missing and not absent, } sources = { "doc-1": "The service retains audit logs for 30 days.", "doc-2": "Enterprise plans can export logs as JSON." } claim = Claim "All plans retain exportable audit logs for 90 days.", "doc-1", "doc-2" , {"all plans", "90 days"}, print verify claim, sources Expected result: ... 'missing terms': '90 days', 'all plans' , 'supported': False} This is intentionally not semantic reasoning. It catches missing references and obvious term mismatches while remaining easy to inspect. That makes it a useful first test and a poor final verifier. 30 and thirty can be compared.Measure precision and recall separately. A gate that blocks every answer has perfect recall for bad answers and no product value. A gate that approves everything is fast but meaningless. The public MonkeyCode repository https://github.com/chaitin/MonkeyCode describes AI task and project-requirement workflows. Citation gates can be useful anywhere an assistant summarizes repository requirements or task evidence, but this exercise does not test MonkeyCode or describe its implementation. Disclosure: I contribute to the MonkeyCode project. The repository description is public; this Python lesson is independent. After this exercise, the key lesson should be clear: a citation is an address, not proof. Verification needs its own explicit step.