Skip to content
Flows
Tier 3 · Agent memory Agent memory · fetched 2026-07-02

Octopus_mem

An agent memory system built on a skill + memory index architecture following a radical-minimalism principle ('jsonl, md, at most sqlite'), with a dual-repo split: open framework, private data.

View source on GitHub

Key takeaways

  • 01

    A memory index file loaded every session is the core mechanism

  • 02

    Minimal formats (jsonl/md/sqlite) keep memory inspectable

  • 03

    Framework-public, data-private is a clean split

Flows built on this research

From the archive

Verbatim excerpts mined from our local archive of this repository — the prompts, schemas, and patterns worth stealing.

Memory

Skill-scoped memory retrieval with time decay

Retrieval strategy:
- Conversation-as-retrieval: every conversation automatically retrieves related memories
- Skill-centric: only retrieve memories related to the orchestrator skill
- Time decay: more recent memories weigh higher (7-day decay)
- Minimal results: return the 5 most relevant memories

Configuration (memory_index_config.json):
- Retrieval params: max_results=5, decay_days=7
- Time decay: mixed linear + exponential decay
- Cache strategy: 300s TTL cache

[translated] A multi-agent memory system where each agent role gets its own skill-scoped index, retrieval is automatic per conversation, and recency decays over 7 days.

skills/memory_molt/SKILL.md