Every method maps to one lifecycle

Setup

Adding Claims

Extraction modes

Querying

Understanding Your Knowledge Base

Health score breakdown

Knowledge Topology

Provenance and Trust

Knowledge Intelligence

Nine methods that answer questions only a provenance-tracking database can answer.

Research Questions

Backup and Restore

LLM Providers

Set the environment variable for your provider, then configure:

No API key? Use "heuristic" mode — it works entirely offline.

Data Types

Adding a claim

db.ingest(
    subject=("name", "type"),          # e.g. ("api-gateway", "service")
    predicate=("relationship", "class"), # e.g. ("depends_on", "depends_on")
    object=("name", "type"),             # e.g. ("redis", "service")
    provenance={
        "source_type": "...",           # What kind of source
        "source_id": "...",             # Identifies the specific source
    },
    confidence=0.9,                    # 0.0 to 1.0 (optional)
    payload={...},                     # Any structured data (optional)
)

Query result

frame = db.query("redis")
frame.focal_entity           # The entity you queried
frame.claim_count            # Number of claims about it
frame.direct_relationships   # Connected entities with predicates and confidence
frame.narrative              # Human-readable summary

Batch input

from attestdb import ClaimInput

claims = [
    ClaimInput(
        subject=("api-gateway", "service"),
        predicate=("depends_on", "depends_on"),
        object=("redis", "service"),
        provenance={"source_type": "config_management", "source_id": "k8s"},
    ),
    # ... more claims
]
result = db.ingest_batch(claims)

Event Hooks

Subscribe to lifecycle events. Callbacks run synchronously after the operation completes. Errors in callbacks are logged, never propagated — your pipeline keeps running.

Events

Example

def on_new_claim(claim_id, claim_input, **kw):
    print(f"New claim: {claim_id}")

def on_corroboration(content_id, count, **kw):
    print(f"Corroborated! {count} independent sources")

db.on("claim_ingested", on_new_claim)
db.on("claim_corroborated", on_corroboration)

Agent Integration

Two ways for external agents to read and write Attest — choose based on your agent framework.

MCP Server (Model Context Protocol)

For Claude Desktop, Claude Code, and any MCP-compatible agent. Ships as a CLI tool.

$ pip install attestdb[mcp]
$ ATTEST_DB_PATH=my.db attest-mcp

Exposes 15 tools (ingest_claim, query_entity, search_entities, knowledge_health, retract_source, etc.) and 2 resources (attest://entities, attest://schema) over stdio transport.

Claude Desktop configuration

{
  "mcpServers": {
    "attest": {
      "command": "attest-mcp",
      "env": {
        "ATTEST_DB_PATH": "my_knowledge.db"
      }
    }
  }
}

REST API

For web-based agents, custom integrations, or any HTTP client. Mounted at /api/v1/ in the Attest Console.

$ pip install attest-console
$ attest-console my.db

Example

# Ingest a claim via REST
curl -X POST http://localhost:8877/api/v1/claims \
  -H "Content-Type: application/json" \
  -d '{
    "subject": ["api-gateway", "service"],
    "predicate": ["depends_on", "dependency"],
    "object": ["redis", "service"],
    "source_type": "k8s_manifest",
    "source_id": "deploy/prod"
  }'

# Query an entity
curl http://localhost:8877/api/v1/entities/redis/context

# Check knowledge health
curl http://localhost:8877/api/v1/health