Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Patrick is a licensed expertise management system built around executive decision-making workflows. It exposes structured expertise templates via a CLI that agents fetch, pass to an LLM for processing, and optionally store back for later retrieval. Expertise is organized into categories like sense, decide, align, and execute.
Unlike raw LLM prompting, Patrick provides versioned expertise templates with JSON schema validation and bidirectional data storage, keeping executive workflows reproducible and auditable.
When to use it
- Generating a daily executive briefing from operational data
- Framing an ambiguous business decision into structured trade-offs
- Scheduling recurring strategic analysis via cronjobs
- Storing and retrieving LLM-processed decision outputs over time
- Running context-aware expertise against Slack archives and JIRA tickets
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Frame the decision about whether to raise prices
- 1Run patrick-cli list to identify available expertise and required context variables
- 2Fetch the decision-framing expertise with context: patrick-cli fetch decision-framing --context '{"decision":"Should we raise prices?"}'
- 3Extract the content and response_schema fields from the JSON output
- 4Send the expertise content to the LLM and validate its response against the schema
- 5Store the validated response back with patrick-cli send decision-framing --data @response.json
Structured decision framing with trade-offs, options, and recommended next steps stored in Patrick's datastore
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.