sqlite-graphrag 1.2.8

Persistent GraphRAG memory for Claude Code, Codex, Cursor, and 27 AI agents — one self-contained ~19 MiB Rust binary, zero daemon. Never re-explain your codebase again. Hybrid retrieval (FTS5 BM25 + cosine similarity + multi-hop graph traversal) surfaces the right memory in milliseconds. Embedding and entity enrichment run as parallel REST calls against your cloud LLM — no fragile headless subprocesses, no ONNX runtime, no model downloads. Soft-delete with full version history, transactional atomic writes, BLAKE3-tracked mutations. OAuth-only: raw API keys ABORT the spawn.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "https://github.com/danilo-aguiar-br/sqlite-graphrag/schemas/remember-dry-run.schema.json",
  "title": "RememberDryRunResponse",
  "description": "GAP-SG-216: response emitted by `sqlite-graphrag remember --dry-run`. A DIFFERENT shape from remember.schema.json, which requires memory_id, version and thirteen other members a run that wrote nothing cannot supply — so the dry run satisfied no published schema at all until v1.2.8. It now reports what the validation actually learned: the parsed graph sizes and every non-canonical entity_type, the same warnings the real path emits for the same input.",
  "type": "object",
  "required": [
    "dry_run", "name", "namespace", "planned_action",
    "entities_parsed", "relationships_parsed", "warnings"
  ],
  "additionalProperties": false,
  "properties": {
    "agent_surface": { "$ref": "https://github.com/danilo-aguiar-br/sqlite-graphrag/schemas/agent-surface.schema.json#/$defs/AgentSurfaceMeta" },
    "dry_run": {
      "const": true,
      "description": "Discriminator. Always true; the real run emits remember.schema.json instead and never carries this member."
    },
    "name": {
      "type": "string",
      "description": "Normalized kebab-case name the write would use, not necessarily the one typed."
    },
    "namespace": {
      "type": "string",
      "description": "Resolved namespace."
    },
    "planned_action": {
      "type": "string",
      "enum": ["would_create", "would_update"],
      "description": "would_update only when the memory exists AND --force-merge was given."
    },
    "entities_parsed": {
      "type": "integer",
      "minimum": 0,
      "description": "Entities the graph payload yielded, after the per-memory limit was applied."
    },
    "relationships_parsed": {
      "type": "integer",
      "minimum": 0,
      "description": "Relationships the graph payload yielded, after the per-memory limit was applied."
    },
    "warnings": {
      "type": "array",
      "items": { "type": "string" },
      "description": "Declared entity_type labels outside the thirteen recommended kinds, phrased as `entity_type 'X' on entity 'Y' is outside the canonical set`. Since v1.2.8 the vocabulary is OPEN, so this is an advisory and the label would be stored as written; under `--strict-entity-types` the same list is what refuses the write with exit 1 instead. Empty rather than absent when every label is canonical, so a consumer never has to distinguish the two."
    },
    "truncated": {
      "type": "boolean",
      "description": "GAP-SG-142: raised by the agent-native surface whenever it removed data from this envelope."
    },
    "count": {
      "type": "integer",
      "minimum": 0,
      "description": "GAP-SG-142: emitted under --count-only, which replaces the payload."
    }
  },
  "examples": [
    {
      "dry_run": true,
      "name": "alice-onboarding-checklist",
      "namespace": "global",
      "planned_action": "would_create",
      "entities_parsed": 2,
      "relationships_parsed": 1,
      "warnings": ["entity_type 'practice' on entity 'alice-praxis' is outside the canonical set"]
    }
  ]
}