Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
Clarified Natural Language vs PDDL capability to fix security flag. Added Implementation Note explaining Conceptual Mode (LLM) vs Formal Mode (PDDL).
What This Skill Does
Applies RALSTP (Recursive Agents and Landmarks Strategic-Tactical Planning), a methodology from a 2024 King's College London PhD thesis, to analyze complex multi-actor problems. Works in two modes: conceptual for natural language problems, and formal for PDDL domain files. Identifies agents, maps dependency chains, measures entanglement, and suggests how to decompose and parallelize work.
Uses a formal planning framework grounded in AI research to surface hidden dependencies and entanglement that informal planning methods typically overlook.
When to use it
- Planning a multi-team product launch with sequenced handoffs
- Analyzing VM migration dependencies across datacenters
- Mapping task order in emergency response coordination
- Identifying parallelizable steps in a software deployment pipeline
- Decomposing cross-functional project workflows with resource contention
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: RALSTP analyze: I need to migrate 1000 VMs from datacentre A to B with minimal downtime
- 1Identifies agents: network team, storage team, compute orchestration, monitoring
- 2Identifies passive objects: VMs, data volumes, network configurations
- 3Maps the dependency graph and detects entanglement between agents sharing resources
- 4Calculates difficulty score based on agent count and entanglement level
- 5Produces strategic phase (high-level ordering) and tactical phase (detailed execution) with parallelization suggestions
Full RALSTP analysis with agent list, dependency graph, difficulty score, landmark chain, and decomposition recommendations including what can run concurrently