@agentlair/llamaindex
v0.0.1
Published
AgentLair behavioral telemetry adapter for LlamaIndex — reserved namespace stub. See https://agentlair.dev
Maintainers
Readme
@agentlair/llamaindex
AgentLair behavioral telemetry adapter for LlamaIndex.
This is a placeholder package reserving the npm namespace @agentlair/llamaindex. v0.0.1 contains a marker module only. v1 ships AAT injection, trust scoring, and behavioral telemetry hooks for LlamaIndex agents and pipelines.
What this will do
When v1 lands, @agentlair/llamaindex will provide:
- AAT injection — attach the Agent Authentication Token to LlamaIndex agent queries
- Behavioral telemetry — emit structured trust events from RAG pipelines and agent steps
- Trust scoring — query the AgentLair API for trust scores on agent tool calls and retrievals
- Pipeline attestation — attest LlamaIndex pipeline runs via SCITT for audit trails
Why this stub exists
As agent runtimes commoditize, integration namespaces become real estate. Reserving @agentlair/llamaindex now means zero migration cost for downstream consumers when the adapter ships.
Usage (v0.0.1 stub)
import { VERSION, STATUS, ADAPTER, AGENTLAIR_URL } from "@agentlair/llamaindex";
// VERSION === "0.0.1"
// STATUS === "reserved"
// ADAPTER === "llamaindex"Reference
- AgentLair: https://agentlair.dev
- LlamaIndex: https://github.com/run-llama/llama_index
- Source: https://github.com/piiiico/agentlair-primitives
License
Apache-2.0
