GitHub hosts HunyuanVideo, Tencent's open-source framework for large-scale video generation models, enabling AI-driven video creation.
AgentFacts
$0.07Starting price
0Popularity
About AgentFacts
AgentFacts is an open-source Python SDK for creating and managing verifiable AI agent metadata. It allows users to generate signed profiles with DIDs (Decentralized Identifiers), capabilities, and provenance for LangChain, Hugging Face, and custom agents. The tool is self-hosted, framework-agnostic, and tamper-proof. AgentFacts enables users to create identities for their agents, leverage different identity standards, and make their agents ready for agentic commerce like UCP (Universal Content Provider) and discoverability standards like A2A (Agent-to-Agent) and MCP-I (Machine-Readable Content Packaging). The tool is designed for developers who need to manage and verify the metadata of AI agents. AgentFacts provides a flexible and secure way to create, manage, and use verifiable AI agent metadata.
GitHub, Inc.
San Francisco, California, US · Founded 2008
- Founders
- Tom Preston-Werner, Chris Wanstrath, PJ Hyett, Scott Chacon
- Founded
- 2008
- Headquarters
- San Francisco, California, US
- Legal status
- Subsidiary of Microsoft (NASDAQ: MSFT)
Key features
- Open-source Python SDK
- Verifiable AI agent metadata
- Signed profiles with DIDs, capabilities, and provenance
- Framework-agnostic and self-hosted
- Tamper-proof
- Create identities for agents
- Leverage different identity standards
- Make agents ready for agentic commerce
Use cases
- Creating verifiable AI agent metadata for LangChain, Hugging Face, or custom agents
- Managing and verifying the metadata of AI agents in agentic commerce
- Implementing discoverability standards like A2A and MCP-I
Pros
- Generates verifiable, tamper-evident profiles for AI agents using Ed25519 signatures and decentralized identifiers (DIDs)
- Supports auto-introspection of popular frameworks like LangChain, LlamaIndex, Hugging Face, and CrewAI
- Provides structured metadata including model details, tools, policies, and provenance in a standardized schema
- Enables offline verification of agent identities and capabilities without requiring external services
- Offers lightweight signing and RFC 8785 canonical JSON for cross-language compatibility
Cons
- Requires Python 3.10 or higher, limiting compatibility with older Python versions
- Initial setup and integration may involve cryptographic key management, which can be complex for non-technical users
- Tamper-evident logs and Merkle proofs add computational overhead during profile generation and verification
Frequently asked questions about AgentFacts
What is AgentFacts and what does it do?
AgentFacts is an open-source Python SDK designed to create and manage verifiable metadata for AI agents. It generates signed profiles containing identity, capabilities, and provenance using decentralized identifiers (DIDs) and cryptographic signatures, ensuring tamper-proof verification.
Who should use AgentFacts?
AgentFacts is intended for developers and organizations that need to establish trust and transparency for their AI agents, particularly those working with frameworks like LangChain, Hugging Face, or custom agents. It suits use cases requiring verifiable agent identities and compliance tracking.
How does AgentFacts verify agent metadata?
AgentFacts uses Ed25519 signatures and did:key decentralized identifiers to sign and verify agent profiles. The verification process checks the cryptographic signature and Merkle log proofs to ensure the metadata has not been tampered with.
What frameworks does AgentFacts support for introspection?
AgentFacts supports auto-introspection for popular agent frameworks such as LangChain, LlamaIndex, Hugging Face Agents, OpenAgents, CrewAI, and AutoGen. It can also work with custom or manual agent configurations.
How do I get started with AgentFacts?
To get started, install the AgentFacts SDK using pip, such as `pip install agentfacts` or with optional extras like `pip install agentfacts[cli]` for CLI tools. Then, create an agent profile by introspecting an existing agent, signing the metadata, and exporting it as a JSON file.
Can AgentFacts be used for agentic commerce or discoverability standards?
Yes, AgentFacts is designed to make AI agents ready for agentic commerce and discoverability standards like UCP (Universal Content Provider), A2A (Agent-to-Agent), and MCP-I (Machine-Readable Content Packaging) by providing verifiable metadata and identity.
