A high-integrity execution engine for Openclaw Skills that runs immutable, pre-validated plans with zero interpretation and strict termination rules.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install plan-executor
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install plan-executor using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Plan Executor?
The Plan Executor is a specialized skill designed to bridge the gap between autonomous planning and reliable execution within the Openclaw Skills ecosystem. It functions as a deterministic operator that only activates when presented with a plan marked as finalized and immutable. By removing all layers of interpretation, the skill ensures that complex sequences are carried out exactly as specified, providing a robust layer of safety and predictability for agent-led workflows.
In environments where Openclaw Skills are used to manage sensitive system operations, the Plan Executor acts as a critical guardrail. It eliminates the risk of an AI agent 'improvising' or deviating from a validated path, making it an essential tool for developers who require high-stakes automation with absolute certainty in outcome and behavior.
Plan Executor Use Cases
- Executing multi-stage system deployments that require strict sequential ordering.
- Running validated data migration scripts where any deviation could result in data loss.
- Implementing automated disaster recovery steps that must be followed without deviation.
- Safeguarding autonomous agent workflows by separating the planning phase from the execution phase.
How Plan Executor Works
- Activation Trigger: The skill monitors for plans containing explicit FINALIZED, EXECUTION-READY, and IMMUTABLE markers.
- Preflight Verification: It performs a structural audit to ensure step indices are contiguous, all targets are declared, and success/failure conditions are explicitly defined.
- Plan Locking: Once verified, the plan state is locked to prevent any mid-execution modifications or external interference.
- Sequential Execution: Steps are executed one by one in ascending numerical order, applying inputs exactly as written using system-io tools.
- Continuous Monitoring: The skill monitors success conditions for each step; any failure or detection of ambiguity results in an immediate, safe halt of the entire process.
- Final Termination: Upon successful completion of the final step, the skill exits execution mode without further commentary.
Plan Executor Setup
To integrate the Plan Executor with your existing Openclaw Skills setup, ensure the system-io tool is authorized for your agent environment. Use the following commands to initialize the execution module:
# Add the plan-executor skill to your environment
openclaw add skill plan-executor
# Validate the installation and tool dependencies
openclaw check plan-executor --verify-io
Plan Executor Data Schema & Taxonomy
The Plan Executor requires a specific JSON or YAML schema within the Markdown file to initiate. The structure must follow the Openclaw Skills standard for immutable plans:
| Attribute | Requirement | Description |
|---|---|---|
status |
FINALIZED | Must be explicitly set to trigger execution |
mutability |
IMMUTABLE | Ensures the plan cannot be changed during the run |
steps |
Array | A finite, ordered list of discrete actions |
index |
Integer | Numeric IDs starting at 1 with no gaps |
success_condition |
String | A clear, non-ambiguous state that defines step completion |
Plan Executor Advanced Features
- Zero-Interpretation Policy: Prevents the agent from inferring intent, ensuring 1:1 parity between the plan and the action.
- Ambiguity Guardrails: Automatically halts execution if vague terms like 'optimal' or 'reasonable' are detected in the instruction set.
- Atomic Failure Handling: Includes a strict termination protocol that prevents cascading errors by stopping the sequence at the first sign of deviation.
- Passive Mode Integration: Designed to remain entirely dormant within Openclaw Skills until all activation criteria are met, conserving resources and preventing accidental triggers.
SKILL.md
Loading
Related Openclaw Skills
KIMI Delegation Skill
A specialized skill that forces an agent to proxy all reasoning and code generation tasks to a KIMI causal language model.
Human-Optimized Frontend Design
A sophisticated design engine that generates quantified, high-fidelity frontend specifications by optimizing for aesthetics, motion, and user experience.
Deliberate Frontend Redesign
A specialized tool for generating bold, unconventional frontend design specifications that break away from standard UI patterns.
Autonomous Skill Orchestrator
A deterministic coordination engine that manages autonomous multi-skill execution loops under strict safety guardrails.
OpenClaw Subagents Configuration
A specialized skill for designing, spawning, and managing autonomous OpenClaw subagents and multi-agent orchestration patterns.
ecto
A robust command-line interface for managing Ghost.io blog content, tags, and site settings via the Admin API.
Featured*
