> ## Documentation Index
> Fetch the complete documentation index at: https://docs.loom.teamecho.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Loom
> The missing 'last mile' of agentic memory — structured, persistent memory for LLM-powered agents.
## What is Loom?
Loom is the missing "last mile" of agentic memory. Rather than just another storage or retrieval layer, Loom is a **decoupled framework** that compiles candidate memories into the **Minimum Sufficient Context** for LLM agents.
While naive retrieval approaches treat memory as a search problem, Loom treats it as an **associative recommendation problem**. Driven by real-world product abstractions from building life-long AI companions, Loom ensures agents truly remember, evolve, and reason without flooding the context window.
Get Loom up and running in minutes.
Understand how the system works end-to-end.
Use Loom in your Python applications.
Integrate via the built-in HTTP server.
## The Cognitive Pipeline
Through its cognitive architecture, Loom achieves this via a robust pipeline:
* **Goal-Aware Context** — Builds the exact context needed based on the agent's current task and token budget — no more, no less.
* **Associative Recall** — Links related memories naturally, but applies strict depth limits to prevent context bloat.
* **Read & Write Loop** — Pulls information from multiple sources and actively saves high-value insights back to memory.
* **Flexible Schemas** — Support for custom schemas, AI-generated schemas, and production-tested templates out-of-the-box.
* **Decoupled & Inspectable** — Plugs into any existing storage (Vector DBs, RAG) while keeping the memory process human-readable and debuggable.
## Schema: A Cognitive Ontology
Unlike standard RAG systems that retrieve noisy, unstructured dialogue chunks, or static markdown file systems that lack active reasoning semantics, a Loom Schema is a **dynamic, self-describing cognitive ontology**. It acts as an executable "file system for cognition" — structuring memory into an inspectable, actively updatable tree.
### Multi-Resolution Intelligence
Loom's Schema is **Hierarchical & Collapsible**, creating a "Variable Focal Length" for the agent's memory:
* **High-Level Abstraction (Low Resolution)** — The agent perceives the Core Essence: not "Alice likes hiking on Sundays" but `Identity -> High-Energy Adventurer`. Lightning-fast, high-dimensional reasoning.
* **Granular Specificity (High Resolution)** — When precision is needed, the schema unfolds recursively from a general persona down to microscopic nuances of a specific habit.
### Hybrid Schema Strategy
Loom provides three strategies that balance human control with machine intuition:
1. **Pre-designed Schemas** — Define the "Skeleton of Truth" manually for absolute precision.
2. **AI-Generated Schemas** — Auto-generate schemas on the fly as new patterns emerge.
3. **NS-Inspired Reference Schemas** — Battle-tested templates from AI Companionship and Social AI (NatureSelect.ai).
Dive into the Schema Call Protocol architecture.
## Benchmark & Results
## Vision: Cognition, Not Just Memory
Memory is a means, not the end. Loom's "last mile" approach isn't just about saving context window tokens — it's about shifting the paradigm from storing raw data to **evolving cognition**.
By distilling noisy interactions into the Minimum Sufficient Context along a structured "skeleton of truth," Loom bridges the gap between simply knowing what a user said and truly understanding who they are.
The ultimate endgame is **hidden state modeling** — moving beyond explicit dialogue logs to capture underlying intents, beliefs, and emotional baselines. This is the path to AI companions and assistants with genuine intelligence.