Description

Add a Claude Code skill for Mem0 Platform that teaches AI agents how to generate correct, working Mem0 integration code eliminating hallucination from outdated training data.

When installed, Claude can set up Mem0 in a project, generate correct API calls for Python and TypeScript, integrate with popular frameworks, and search live docs on demand all verified against the actual codebase.

Structure (11 files):

• SKILL.md — Lean entry point (~156 lines) with quickstart, core pattern, edge cases
• references/ — 7 on-demand docs:
  • quickstart.md — Python, TypeScript, cURL setup
  • sdk-guide.md — All SDK methods for both languages
  • api-reference.md — REST endpoints, filters, memory object schema
  • architecture.md — Processing pipeline, memory lifecycle, scoping model, performance
  • features.md — Advanced retrieval, graph memory, categories, MCP, webhooks, multimodal, feedback, export
  • integration-patterns.md — LangChain, CrewAI, Vercel AI, OpenAI Agents, LangGraph, LlamaIndex, AutoGen, Pipecat
  • use-cases.md — 7 real-world patterns with full Python + TypeScript implementations
• scripts/mem0_doc_search.py — Live doc search against docs.mem0.ai for staying current

Key decisions:

• Platform-only — all examples use MemoryClient, zero OSS references
• Follows the Agent Skills spec.
• All code verified against mem0/client/main.py, mem0-ts/src/client/mem0.ts, and docs/openapi.json

Type of change

• New feature (non-breaking change which adds functionality)

How Has This Been Tested?

• All SKILL.md links verified — zero broken references
• All API endpoints verified against SDK source (mem0/client/main.py) and OpenAPI spec
• All code examples verified against docs/integrations/*.mdx and docs/platform/*.mdx
• TypeScript SDK methods verified against mem0-ts/src/client/mem0.ts
• Skill structure validated against Agent Skills specification
• Pre-commit hooks pass (Ruff, isort)

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code