Metadata-Version: 2.4
Name: calera-agent-memory
Version: 0.4.0
Summary: Plug-and-play topological memory substrate for LLMs, LangChain, and autonomous agent swarms.
Author-email: Calera Labs <support@caleralabs.com>
License: Apache-2.0
Project-URL: Homepage, https://icx.caleralabs.com
Project-URL: Documentation, https://icx.caleralabs.com/docs
Project-URL: Repository, https://github.com/Calera-Labs/icx-mcp
Keywords: ai-agent,mcp,memory,topological-memory,langchain,crewai,infinite-context
Classifier: Programming Language :: Python :: 3
Classifier: Operating System :: OS Independent
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Requires-Python: >=3.9
Description-Content-Type: text/markdown

# Calera Agent Memory (`calera-agent-memory`)

> The plug-and-play **AI Memory Substrate** for Python agent swarms, LangChain, and CrewAI.  
> Eliminates LLM context rot with sub-5ms $O(1)$ topological associative memory.

---

## ⚡ Quickstart: 3 Lines of Code

```python
from calera_agent_memory import CaleraMemorySubstrate

# Auto-provisions a free 1,000,000-node persistent vault under Delaware UETA § 14
memory = CaleraMemorySubstrate(principal_email="developer@company.com")

# Drop directly into your agent or LLM chain
memory.save_context({"input": "Project Alpha deadline"}, {"output": "October 15, 2026"})
context = memory.load_memory_variables({"input": "When is the project due?"})
print(context["history"])
```

---

## Direct Client Usage

```python
from calera_agent_memory import CaleraMemoryClient

client = CaleraMemoryClient(api_key="your_optional_api_key")

# Claim free vault
vault_info = client.claim_free_vault("me@example.com", agent_name="my-researcher")

# Store fact into 4D topological lattice
client.store("NVDA_FY26_Rev", "Estimated at $168B based on datacenter expansion")

# Sub-5ms recall with zero token bloat
results = client.query("NVDA revenue projection")
print(results)
```

---

## Key Advantages
- **Sub-5ms Recall:** $O(1)$ topological graph retrieval instead of slow vector database linear scans.
- **Zero Hallucination Context:** Retains exact structured relationships without semantic degradation.
- **Free 1M Node Tier:** Agents can self-provision permanent memory vaults at runtime for $0.00.
