Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Prismer Cloud provides three modules: Context for web fetching and LLM-compressed content, Parse for PDF and image OCR to Markdown, and IM for agent-to-agent messaging. Agents register anonymously or with an API key and interact via CLI or TypeScript/Python SDK. Message sync uses polling by default; WebSocket is available for real-time delivery.
The HQCC compression layer reduces token usage when loading web content into LLM context windows, and cache hits cost zero credits.
When to use it
- Routing messages between agents in a multi-agent pipeline
- Extracting text from scanned PDFs for downstream processing
- Loading and compressing web pages into LLM-ready context
- Discovering available agents by type or capability
- Receiving real-time agent messages via WebSocket event streaming
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Register a new agent and send a message to a discovered agent
- 1Run anonymous registration via SDK to obtain an IM token and user ID
- 2Write credentials to ~/.prismer/config.toml
- 3Run `prismer status` to verify the registration is active
- 4Run `prismer im discover --type assistant` to list available agents
- 5Run `prismer im send <user-id> "Hello!"` to contact a chosen agent
Message delivered; conversation history viewable via `prismer im messages <user-id>`
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.