Integration Guide

Signet + LlamaIndex

Integrate Signet trust scoring with LlamaIndex data agents to verify trust in RAG pipelines and data source interactions.

Why integrate Signet with LlamaIndex?

LlamaIndex powers data-aware AI agents that query, retrieve, and synthesize information from diverse data sources. Signet integration adds trust verification for agents in multi-agent retrieval architectures, ensuring data sources and retrieval agents meet trust requirements.

Integration steps

Step 1: Register data agents

Register your LlamaIndex query engines and data agents with Signet to establish trust identities.

Step 2: Verify retrieval sources

Before accepting retrieved data from external agents, check their Signet Scores to ensure data quality.

Step 3: Add trust context to queries

Include trust metadata in your query engine context so responses can be weighted by source agent trust.

Step 4: Build trust through usage

Report successful retrieval and synthesis outcomes to Signet to improve agent scores over time.

Code example

from signet import SignetClient

signet = SignetClient(api_key=os.environ["SIGNET_API_KEY"])

# Verify data source agent trust before retrieval
async def trusted_retrieve(agent_sid: str, query: str):
    score = signet.lookup_score(agent_sid)
    if score["recommendation"] == "Caution":
        raise ValueError(f"Data agent {agent_sid} has low trust: {score['composite_score']}")
    # Proceed with retrieval
    return await query_engine.aquery(query)

Ready to integrate?

Apply for API access to start building trust-verified agent interactions with LlamaIndex.