RFS-1: Agentic Storage SDK
Concept
Build the foundational storage toolkit that any AI agent, regardless of framework, runtime, or chain of origin, can use to store data on Filecoin autonomously.
This includes agent-consumable SDKs, wallet tooling, and drop-in backend adapters that make Filecoin the default storage layer for agent-native applications. When an agent needs to persist files, memory, artifacts, or platform data, it should reach for Filecoin, not S3.
The SDK should support:
- Uploading and managing data on Filecoin via FOC's warm storage service, with continuous onchain payment and PDP verification
- Managing storage lifecycle autonomously (uploading pieces to datasets, monitoring proof status, adjusting payment rates)
- Paying for storage continuously without human involvement
- Plugging into agent-native platforms as a drop-in storage backend (replacing S3, centralized databases, and human-owned infra accounts)
What You Might Build
- Agent-oriented wallet tooling (built or integrated) enabling autonomous payment
- Framework-agnostic SDK with clear integration guides, targeting latest FOC SDK version
- Platform adapters usable by agent runtimes (compatible with social, coordination, and publishing networks)
- Default storage policies tuned for agents (data redundancy, proof frequency, payment rate)
- Deployed FOC-backed storage on calibnet (extra point for mainnet), with transparent cost and lifecycle management
Why This Matters
This is the canonical storage primitive for AI agents on Filecoin. Every other RFS in this set depends on agents being able to store data cheaply, verifiably, and without human intermediation. This SDK is the entry point.
What This Could Become
- Agent memory platform. A hosted service where agents from any framework store and retrieve long-term memory - conversation history, learned preferences, accumulated knowledge - persisted on Filecoin. Agents switch runtimes or providers without losing their memory. Think "Mem0 but decentralized."
- Agent-native backup and disaster recovery. Autonomous agents that continuously snapshot their own state to Filecoin, with rollback capabilities. If an agent crashes or gets corrupted, it restores from its last verified checkpoint - agents that survive infrastructure failures without human intervention.
- Drop-in S3 replacement for AI-native apps. An API-compatible storage layer that any AI application can swap in for S3, with verifiable persistence and no vendor lock-in. Target the wave of AI-native startups that don't want AWS dependency.
How to Get Started
- Define a minimal agent storage interface (upload, retrieve, monitor proofs, manage payment)
- Build or integrate wallet tooling suitable for agent use (evaluate ERC-4337 or custom)
- Build adapters for at least two agent-native platforms or frameworks
- Deploy on FOC calibnet and mainnet with default storage policies
- Publish agent-consumable documentation (e.g., skills.md or equivalent)
Key Links
A Note on FOC's Storage Model
FOC uses PDP-based (Proof of Data Possession) warm storage with continuous onchain payment rails. You upload pieces to datasets, they get proven every period, and you pay an ongoing rate. This is not the traditional Filecoin storage deal model based on PoRep (Proof of Replication). Think of PDP as: upload data, it stays warm and provably available, you pay continuously.
SDK Maturity & Builder Expectations
The FOC SDK is under active development. There are known issues pending upgrade. Mainnet-ready milestone (M4.1) is targeted for approximately March 14, 2026. Builders should:
- Target the latest SDK version once available; earlier versions may have breaking changes
- Expect API instability - interfaces may shift before M4.1
- Reach out to the FOC team when things break; active support is available during the build period
We'll update this page as the SDK stabilizes.