Agent Skills as Memory Layer

Skill Memory for
Claude Code

Read SKILL.md to set up Acontext for Claude CodeCopied!

With Acontext, your agents can improve themselves via dynamic skill.

Session BackupImprove Your Agents over TimeSync Skills to Local

Captures what works, distills it into skills, and sync them to local.

How Acontext Connects

Your agent runs, Acontext captures the session, distills skills, and feeds them back — a continuous learning loop.

AI Agents
Acontext
Skills Reuse
next run

How Skill Memory Works

Learn from runs. Write as Markdown. Reuse anywhere. Skill memory that turns what your agents did into files they can read and use again.

Session
Extracted Tasks
Waiting for tasks...

Get Started in Minutes

An Acontext API key and pip install acontext — that's all you need.

Skill Memory
skill_memory.py
from acontext import AcontextClient

client = AcontextClient(api_key="sk-ac-...")

# Create a learning space and attach a session
space = client.learning_spaces.create()
session = client.sessions.create()
client.learning_spaces.learn(space.id, session_id=session.id)

# Run your agent, store messages — when tasks complete, learning runs automatically
client.sessions.store_message(session.id, blob={"role": "user", "content": "My name is Gus"})
client.sessions.store_message(session.id, blob={"role": "assistant", "content": "Hi Gus! How can I help you today?"})
# ... agent runs ...

# List learned skills (Markdown files)
client.learning_spaces.wait_for_learning(space.id, session_id=session.id)
skills = client.learning_spaces.list_skills(space.id)

# Download all skill files to a local directory
for skill in skills:
    client.skills.download(skill_id=skill.id, path=f"./skills/{skill.name}")

Create a learning space, attach a session, and Acontext builds skill files from successful runs.

Docs

Supporting Skill Memory

What we offer: skill memory — and the SDKs to use it.

Skill Memory

Learns from task outcomes → writes Markdown files (SKILL.md schema) → agent recalls via get_skill / get_skill_file. Human-readable, portable, no embeddings. Export as ZIP, use in any framework.

SDKs & Integrations

Ready to use with OpenAI, Anthropic, LangGraph, Agno, and other popular agent frameworks.

All of this exists to feed one thing: a skill memory layer you can read, edit, and move.

Open-source, one command to host

Self-Host in Seconds

Run the full Acontext stack on your own infrastructure with a single command.

curl -fsSL https://install.acontext.io | shCopied!

Join the Community

Connect with early builders & preview new features