signetai

**Store, sync and share memories, system prompts, transcripts, institutional knowledge, and secrets between all of your favorite harnesses and models.** **97.6% average LongMemEval answer accuracy** [Quick start](https://docs.signetai.sh/quickstart/) · [Why Signet](#why-signet) · [Benchmarks](https://docs.signetai.sh/benchmarking/) · [Docs](https://docs.signetai.sh/quickstart/) · [Discord](https://discord.gg/Psdeg7sQm7)

Signet automatically creates memories from your transcripts, imported files, and other sources. Memories are built and maintained in the background by a process called dreaming, which constructs a living semantic ontology—a structured representation of your world—on top of your raw history. This turns an agent's transcripts and data into connected memory with an audit trail back to the source.

The result: the agent gets the right context before the next prompt starts, with a path back to the raw source when deeper context is needed.

This is useful for:

• Company brains—Connect Signet to all the tools in your existing stack, and watch your agents spend less time learning your business and more time being useful to it.
• Developers—Managing projects across different models and harnesses usually means context gets fragmented across each one. With Signet, ChatGPT, Claude, Hermes, and Pi can all work off the same shared knowledge base. Your agents suddenly know what's going on, and how they fit into the bigger picture.
• Individuals—Research, journaling, daily work: run the same agent across all of it without re-explaining yourself every session. History compounds instead of resetting.
• Autonomous agents—Scheduled agents that run unattended, like a morning-brief or monitoring agent, keep continuity between runs without a human re-priming them each time.
• Agent builders—Ship an agent product without building memory infrastructure from scratch. Signet is the memory layer underneath, with the audit trail doubling as a debugging and trust feature.

Read more: Why Signet · Architecture · Knowledge Graph · Pipeline

Quick start (about 5 minutes)

Install Signet

curl -fsSL https://signetai.sh/install.sh | bash                 # recommended

Or: npm install -g signetai / bun add -g signetai

Don't want to handle setup yourself? Paste this to your AI agent:

Install and fully configure Signet AI by following this guide exactly: https://signetai.sh/skill.md

Covers Linux x64/arm64, macOS x64/arm64, Docker, and Windows x64 (Windows: use npm install -g signetai or Docker).

Setup

signet setup                         # interactive setup wizard
signet status                        # confirm daemon + pipeline health
signet dashboard                     # open memory + retrieval inspector

Harness support

Signet runs underneath the tools you already use. Run signet setup to configure plugins and connectors. Currently, Signet supports:

Don't see your favorite harness? File an issue and request that it be added!

Signet supports a wide variety of sources that can be imported directly into your agent's memory graph — included in dreaming sessions and surfaced as new connections in recall.

Supported formats for one-time import:

Documentation

• Quickstart
• CLI Reference
• Configuration
• Telemetry
• Hooks
• Harnesses
• Secrets
• Skills
• Auth
• Dashboard
• SDK
• API Reference
• Knowledge Architecture
• Knowledge Graph
• Benchmarks
• Roadmap
• Repository Map

Benchmarks

Signet's latest tracked MemoryBench run averages 97.6% LongMemEval answer accuracy.

The benchmark matters because local custody should not mean weak recall. Signet is designed to retrieve the right facts across long-running, multi-session conversations while keeping memory inspectable and repairable.

See Benchmarks for the methodology, scoring note, and run workflow.

Development

git clone https://github.com/Signet-AI/signetai.git
cd signetai

bun install
bun run build
bun test
bun run lint

cd platform/daemon && bun run dev     # Daemon dev (watch mode)
cd surfaces/dashboard && bun run dev  # Dashboard dev

Requirements:

• Bun for normal repo development
• Node.js 18+ for Node-targeted package surfaces
• macOS or Linux
• Optional for harness integrations: Claude Code, Codex, Kimi Code, OpenCode, OpenClaw, Gemini CLI, Pi, Oh My Pi, or Hermes Agent

Contributing

New to open source? Start with Your First PR. For code conventions and project structure, see CONTRIBUTING.md. Open an issue before contributing significant features. Read the AI Policy before submitting AI-assisted work.

Star History

Contributors

Made with love by...

License

Apache-2.0.

signetai.sh · docs · spec · discussions · issues