Why Agents Need a Passport: .fafa — Portable Identity for the Agentic Era

This paper specifies .fafa (application/vnd.fafa+yaml), the IANA-registered media type for declarative agent identity: a portable passport for who an agent is, what it may do, where it is reached, and what it must never do. It complements Agent Cards, Model Context Protocol server-cards, and AGENTS.md; it does not replace them.

In production, that identity is usually inferred from system prompts, product settings, and protocol-native cards that do not travel cleanly across hosts — seam rewrites, inference cost, and trust surfaces that cannot be diffed. A .fafa document is simple, structured, and persistent. Its lifecycle is author → serve → discover → project. House cards stay the wire format; AGENTS.md stays the repo ops briefing; .fafa answers who the agent is. The paper is self-contained. Multi-agent orchestration, deep composition, and scoring are deferred.

faf-agent-passport-v1.0.pdf

Files (462.1 kB)

• IANA. Media type application/vnd.fafa+yaml. Registered 2026-06-26. https://www.iana.org/assignments/media-types/application/vnd.fafa+yaml
• IANA. Media type application/vnd.faf+yaml. Registered 2025-10-31. https://www.iana.org/assignments/media-types/application/vnd.faf+yaml
• IANA. Media type application/vnd.fafm+yaml. Registered 2026-05-13. https://www.iana.org/assignments/media-types/application/vnd.fafm+yaml
• AGENTS.md. Repository-level agent instructions. https://agents.md
• Model Context Protocol (MCP). https://modelcontextprotocol.io
• Agent-to-Agent (A2A). Agent Card discovery and interoperability.
• Polli, R., Wilde, E., and Aro, E. YAML Media Type. RFC 9512, February 2024. https://www.rfc-editor.org/rfc/rfc9512
• Wolfe, J. Format-Driven AI Context Architecture: The .faf Standard for Persistent Project Understanding. Zenodo. https://doi.org/10.5281/zenodo.18251362
• Wolfe, J. Permanent Memory and Instant Recall: The .fafm Standard for Multi-Profile AI Agent Memory. Zenodo. https://doi.org/10.5281/zenodo.20348942
• Agentic AI Foundation (AAIF). https://aaif.io