1{
2  "$schema": "https://json-schema.org/draft/2020-12/schema",
3  "title": "Evidence",
4  "description": "A piece of evidence extracted from a source. evidence_id = sha256(source_id + normalized_quote + locator)[:16].",
5  "type": "object",
6  "required": ["evidence_id", "source_id", "quote", "evidence_type", "captured_at"],
7  "properties": {
8    "evidence_id": {
9      "type": "string",
10      "pattern": "^[0-9a-f]{16}$",
11      "description": "sha256(source_id + normalized_quote + locator)[:16]"
12    },
13    "source_id": {
14      "type": "string",
15      "pattern": "^[0-9a-f]{16}$",
16      "description": "References a source in sources.jsonl"
17    },
18    "retrieval_query": {
19      "type": ["string", "null"],
20      "description": "The search query or prompt that led to this evidence",
21      "default": null
22    },
23    "locator": {
24      "type": ["string", "null"],
25      "description": "Page number, section heading, URL fragment, or timestamp within the source",
26      "default": null
27    },
28    "quote": {
29      "type": "string",
30      "description": "Exact or near-exact text extracted from the source"
31    },
32    "evidence_type": {
33      "type": "string",
34      "enum": ["direct_quote", "paraphrase", "data_point", "figure_reference", "methodology"],
35      "description": "How the evidence was captured"
36    },
37    "captured_at": {
38      "type": "string",
39      "format": "date-time"
40    }
41  },
42  "additionalProperties": false
43}
44