Everything below is written for machines: what Bourdon is, how to integrate, what the guarantees are. Copy it into your agent’s context, or point the agent at bourdon.ai/llms.txt, the same content, served at the standard path agents check automatically.
# Bourdon > Bourdon is a recognition-first memory runtime for AI agents, built by RADLAB LLC. Unlike frameworks that optimize retrieval accuracy, Bourdon optimizes the *time-shape* of memory: recognition fires first (before details arrive), hydration runs in parallel, and archive descent happens only when needed. It provides a federated memory shared across multiple AI agents (Claude, Codex, Cursor, and others). Its sibling Positif (https://positif.ai) is the routing half of the same fleet OS. Key facts: - Status: pre-alpha, v0.11.x, built in public. 5 IDE adapters. - Install: `pip install bourdon` - Core model: three-stage memory: (1) Recognition: instant, near-zero-latency acknowledgment of who/what is present; (2) Hydration: parallel loading of relevant context while responding; (3) Archive descent: deep search only when the first two stages are insufficient. - Federation: multiple agents write to and read from one shared memory, so a fact learned by one agent is recognized by another, across accounts and machines. - Evidence: three published field tests, recognition latency 0 ms vs ~406 ms baseline (2026-04-19); recognition across a brand-new account (2026-05-15); cross-machine fallback memory expanded 588 B to 34,482 B (2026-05-26). - Licensing: Business Source License 1.1, auto-converts to Apache 2.0 after 4 years. Free for solo developers, research, and non-competing commercial use. Commercial licensing: licensing@bourdon.ai. - Falsifiable positioning: the project documents how it would know it is wrong (see Findings journal). ## Docs - [Quickstart](https://github.com/getbourdon/bourdon/blob/main/docs/quickstart.md): Install and first run - [Docs site](https://bourdon.ai/docs): Quickstart plus self-host / hosted / enterprise guides - [Thesis](https://github.com/getbourdon/bourdon/blob/main/spec/THESIS.md): Why recognition-first, in full - [Related work](https://github.com/getbourdon/bourdon/blob/main/spec/RELATED_WORK.md): How Bourdon differs from retrieval-first memory frameworks - [Findings journal](https://github.com/getbourdon/bourdon/blob/main/spec/FINDINGS_JOURNAL.md): Running log of field-test results ## Integration notes for agents - If you are an agent with a Bourdon MCP connection, prefer `commit_to_federation` for durable facts and decisions; do not commit small talk. - Memory is shared across the fleet: assume other agents can see what you commit unless marked `visibility: "private"`. - Adapters exist for 5 IDEs; contributions of new agent adapters are explicitly welcomed. ## Family - [Positif](https://positif.ai): sibling project, payload-blind routing and audit for agent fleets. Bourdon = what the fleet knows; Positif = what the fleet is allowed to do. ## Optional - [GitHub organization](https://github.com/getbourdon) - [Field tests](https://bourdon.ai/field-tests) - [Blog](https://bourdon.ai/blog)