---

name: cli-and-cloud description: "Routes Memori Cloud, CLI, quota/sign-up/setup, MCP, and agent API workflows." metadata: disco-role: operating disable-model-invocation: true license: NOASSERTION

CLI and Cloud

Use this sub-skill for Memori Cloud setup, the python -m memori command, quota and sign-up flows, MCP wiring, and agent-memory API calls.

Use when

• The request mentions Memori Cloud, MEMORI_API_KEY, python -m memori, quota, sign-up, setup, MCP, Claude Code, Hermes, OpenClaw, or agent recall.
• The user is debugging cloud auth, quota, SSL, timeout, or payload-validation failures.
• The task is about Cloud-facing Python or TypeScript quickstarts, not BYODB storage or LLM provider registration.

Read first

• references/cli-reference.md for commands and flags.
• references/cloud-agent-api.md for cloud agent methods and parameter rules.
• references/mcp-and-agent-integrations.md for MCP and adjacent agent integrations.
• references/troubleshooting.md for cloud/auth/network failures.
• scripts/check_memori_cli.py for a safe CLI smoke.

What this sub-skill owns

• Cloud mode vs BYODB mode selection when conn is omitted.
• API key and base URL handling.
• CLI command routing and .env loading from the current directory.
• Cloud agent recall, recall-summary, compaction, capture-turn, and feedback.
• MCP connection headers and agent integration setup notes.

What it does not own

• Storage schema or driver setup: use byodb-storage.
• llm.register(...) provider selection: use llm-integration.
• Recall, embeddings, session lifecycle, or native Rust-core behavior: use memory-and-search.
• TypeScript implementation details: use typescript-sdk.

Safe first check

Run the bundled CLI smoke before suggesting a live cloud call:

python scripts/check_memori_cli.py

That helper only exercises the installed python -m memori entry path from a fresh temporary working directory.