Bỏ qua tới nội dung
TakoAPI
Quay lại kỹ năng

hub1

bởi topguyaii

DevOps & Cloud
—Chưa có đánh giá
Đánh giá của bạn

Tóm tắt

**Memory Infrastructure for Autonomous Agents**.

Mô tả

Kỹ năng này làm gì

OpenClawdy provides persistent memory storage for AI agents across sessions. Agents store facts, preferences, decisions, and learnings, then retrieve them via semantic search. Advanced features include reputation scoring for proven strategies, cross-agent shared memory pools, and snapshot-based state restoration.

Semantic recall retrieves memories by meaning rather than exact keywords, so agents surface relevant context even when query phrasing differs from what was originally stored.

Khi nào nên dùng

  • Remembering user coding preferences across separate sessions
  • Sharing bug discoveries and fixes between agents on a team
  • Rolling back agent memory after an experiment goes wrong
  • Ranking error-handling strategies by past success rate
  • Carrying project context into a new conversation without re-explaining

Quy trình ví dụ

ĐẦU VÀO

User asks: Create a new API endpoint for user authentication

AGENT
  1. 1

    Calls memory_recall to retrieve stored preferences about coding style and tech stack

  2. 2

    Finds preference memory: 'User prefers TypeScript for all code'

  3. 3

    Finds fact memory: 'Project uses Next.js 14 with PostgreSQL'

  4. 4

    Generates a TypeScript endpoint consistent with stored preferences

  5. 5

    Calls memory_store to record the architectural decision made during this session

ĐẦU RA

TypeScript API endpoint code, with the agent citing stored preferences as the basis for its implementation choices

Yêu cầu

  • Agent wallet address configured for wallet-based authentication

Cài đặt

clawhub install topguyaii/hub1

Chạy lệnh CLI, hoặc dán prompt vào Claude Code, Cursor, Codex hay coding agent bạn đang dùng.

Liên kết