Agents & the cachly Brain

Autonomous AI agents (LangChain, CrewAI, AutoGen, custom REST loops) can use the cachly Brain MCP server to persist memory across runs — so every agent instance learns from past attempts and never repeats the same mistakes.

Why do agents need persistent memory?

No re-research

Agent recalls past solutions in <10 ms — skips re-running searches, API calls, or reasoning chains.

Cross-run learning

Lessons from run #1 are available in run #2, across machines and instances.

Fewer failures

Known pitfalls are stored. Agent avoids repeating failed approaches automatically.

Setup — one command

# Detects your editor and writes all configs automatically
npx @cachly-dev/mcp-server@latest autopilot

Or add manually to your agent's MCP config:

{
  "mcpServers": {
    "cachly": {
      "command": "npx",
      "args": ["-y", "@cachly-dev/mcp-server@latest"],
      "env": {
        "CACHLY_JWT": "your-api-key",
        "CACHLY_BRAIN_INSTANCE_ID": "your-instance-uuid"
      }
    }
  }
}

Python agent with Brain memory

import anthropic, os

client = anthropic.Anthropic()

# Brain MCP server — memory persists across every run
server_params = {
    "command": "npx",
    "args": ["-y", "@cachly-dev/mcp-server@latest"],
    "env": {
        "CACHLY_JWT": os.environ["CACHLY_JWT"],
        "CACHLY_BRAIN_INSTANCE_ID": os.environ["CACHLY_BRAIN_INSTANCE_ID"],
    },
}

with client.beta.messages.stream(
    model="claude-opus-4-5",
    max_tokens=4096,
    tools=[{"type": "mcp", "server": server_params}],
    messages=[{
        "role": "user",
        "content": "Fix the flaky Stripe webhook test. Check what worked before first.",
    }],
) as stream:
    for text in stream.text_stream:
        print(text, end="", flush=True)

# The agent automatically:
# 1. Calls recall_best_solution("fix:stripe-webhook") before writing code
# 2. Calls learn_from_attempts(...) after the fix — stored forever

Core Brain tools for agents

recall_best_solution(topic)Most impactful

Before attempting a task, retrieve the best known solution. Returns what worked, what failed, exact commands, and severity.

learn_from_attempts(...)After every run

Store what the agent learned — outcome, approach, commands, file paths. Retrieved by all future agent runs automatically.

smart_recall(query)Semantic search

Natural language search over all stored lessons. E.g. 'docker healthcheck IPv6 error' returns the relevant fix instantly.

remember_context(key, value)State

Persist arbitrary agent state between runs. E.g. last processed record ID, config values, intermediate results.

recall_context(key)State

Retrieve stored context by key. Supports glob patterns (e.g. file:*) for namespace retrieval.

Framework compatibility

Claude Code / Claude APINative MCP

LangChainMCP tool adapter

CrewAIMCP tool adapter

AutoGenMCP tool adapter

OpenAI Agents SDKMCP tool adapter

Google ADKMCP tool adapter

Get Started Free

Free tier included · No credit card