Metadata-Version: 2.4
Name: llamaindex_agentmesh
Version: 3.3.0
Summary: AgentMesh trust layer integration for LlamaIndex agents
Author: AgentMesh Contributors
License: MIT
License-File: LICENSE
Requires-Python: <4.0,>=3.9
Requires-Dist: cryptography<47.0,>=45.0.3
Requires-Dist: llama-index-core<0.15.0,>=0.13.0
Description-Content-Type: text/markdown

# LlamaIndex AgentMesh IntegrationAgentMesh trust layer integration for LlamaIndex - enabling cryptographic identity verification and trust-gated agent workflows.## OverviewThis integration provides:- **TrustedAgentWorker**: Agent worker with cryptographic identity and trust verification- **TrustGatedQueryEngine**: Query engines with access control based on trust- **Secure Data Access**: Governance layer for RAG pipelines with identity-based policies## Installation```bashpip install llama-index-agent-agentmesh```## Quick Start### Creating a Trusted Agent```pythonfrom llama_index.agent.agentmesh import TrustedAgentWorker, VerificationIdentity# Generate cryptographic identityidentity = VerificationIdentity.generate(    agent_name="research-agent",    capabilities=["document_search", "summarization"],)# Create trusted agent workerworker = TrustedAgentWorker.from_tools(    tools=[search_tool, summarize_tool],    identity=identity,    llm=llm,)# Create agent with trust verificationagent = worker.as_agent()```### Trust-Gated Query Engine```pythonfrom llama_index.agent.agentmesh import TrustGatedQueryEngine, TrustPolicy# Wrap query engine with trust policytrusted_engine = TrustGatedQueryEngine(    query_engine=base_engine,    policy=TrustPolicy(        min_trust_score=0.8,        required_capabilities=["document_access"],        audit_queries=True,    ),)# Query requires verified identityresponse = trusted_engine.query(    "What are the quarterly results?",    invoker_card=requester_card,)```### Multi-Agent Trust Handoffs```pythonfrom llama_index.agent.agentmesh import TrustHandshake, TrustedAgentCard# Create agent card for discoverycard = TrustedAgentCard(    name="research-agent",    description="Performs document research",    capabilities=["search", "summarize"],    identity=identity,)card.sign(identity)# Verify peer before task handoffhandshake = TrustHandshake(my_identity=identity)result = handshake.verify_peer(peer_card)if result.trusted:    # Safe to delegate task    pass```## Features### TrustedAgentWorkerAn agent worker that:- Has cryptographic identity for authentication- Verifies peer agents before accepting tasks- Signs outputs for verification by recipients- Supports capability-based access control### TrustGatedQueryEngineA query engine wrapper that:- Requires identity verification for queries- Enforces trust score thresholds- Restricts access based on capabilities- Provides audit logging of all queries### Data Access GovernanceControl access to your RAG pipeline:```pythonfrom llama_index.agent.agentmesh import DataAccessPolicypolicy = DataAccessPolicy(    allowed_collections=["public", "internal"],    denied_collections=["confidential"],    require_audit=True,    max_results_per_query=100,)# Apply policy to indextrusted_index = TrustedVectorStoreIndex(    index=base_index,    policy=policy,)```## Security ModelAgentMesh uses Ed25519 cryptography for:- **Identity Generation**: Unique DID per agent- **Request Signing**: All queries are signed- **Response Verification**: Outputs can be verified## API Reference| Class                   | Description                          || ----------------------- | ------------------------------------ || `VerificationIdentity`          | Cryptographic agent identity         || `TrustedAgentWorker`    | Agent worker with trust verification || `TrustGatedQueryEngine` | Query engine with access control     || `TrustHandshake`        | Peer verification protocol           || `TrustedAgentCard`      | Agent discovery card                 || `DataAccessPolicy`      | RAG access governance                |## LicenseMIT License