A state-driven execution framework that manages task tracking and session recovery for AI coding agents through structured markdown documentation.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install spec-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 spec-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 Spec Executor?
Spec Executor is a specialized execution companion designed to work alongside planning workflows within the Openclaw Skills ecosystem. Its primary purpose is to provide state assurance and navigation during complex coding tasks, ensuring that the AI agent always knows its current position, what comes next, and how to recover if the session is interrupted. By relying on a structured tasks.md file, it transforms abstract plans into incremental, verifiable code deliveries.
This skill is built for developers who need high-precision AI assistance. It enforces a strict discipline of updating state before and after every code change, preventing the AI from drifting away from the original requirements. As a robust implementation of Openclaw Skills, it bridges the gap between high-level architectural design and the granular reality of file-system modifications.
Spec Executor Use Cases
- Managing multi-file implementation tasks where context retention is critical.
- Ensuring code quality through mandatory quality gates and pre-commit checklists.
- Recovering agent state automatically after context window compression or session resets.
- Implementing features incrementally to allow for granular human review and approval.
How Spec Executor Works
- The skill validates that a tasks.md file exists, generated by a preceding planning phase.
- It classifies the current task to determine if it can be executed directly or requires further architectural planning.
- Before editing code, it updates the tracker to mark the specific task as in-progress to prevent duplicate work.
- It implements the code change within a strict limit of 300 lines to ensure reviewability.
- The skill runs automated verification tools like compilers or grep-based signature checks to validate the work.
- It records the verification results and commit message in the tracker before moving to the next task.
Spec Executor Setup
Spec Executor requires a tasks.md file produced by the spec-workflow skill. To initialize the environment for this skill, ensure your project documentation is organized:
# Create a documentation directory for your feature
mkdir -p docs/your-feature-name/
# Ensure spec-workflow has generated the implementation plan
cat docs/your-feature-name/tasks.md
The agent will then automatically recognize the tracker and begin the state-navigation process.
Spec Executor Data Schema & Taxonomy
The skill relies on a standardized Markdown schema for state tracking. The tasks.md file serves as the single source of truth:
| Element | Description |
|---|---|
| Status Tags | [ ] Pending, [~] In-Progress, [\u2713] Done, [\u23ed] Skipped. |
| Scope | Granular definition of files, methods, and fields affected by the task. |
| Verification | Detailed steps and commands used to verify task completion. |
| Meta Tracking | Incremental counters for user corrections and session recoveries to monitor workflow health. |
Spec Executor Advanced Features
- Session Recovery Protocol: Automatically identifies the boundary between completed and pending work after a context reset.
- Continuous Mode Authorization: Allows for batch execution of multiple tasks when explicit user permission is granted.
- Pre-flight Risk Prediction: Analyzes task types (e.g., API alignment, DTO addition) to identify common failure patterns before writing code.
- Tool-Based Validation: Enforces the use of bash-based verification (grep, diff, build tools) over unreliable model-based self-checking.
SKILL.md
Loading
Related Openclaw Skills
Business Blueprint Skill
A technical tool for converting presales requirements and meeting notes into structured business capability blueprints and architectural diagrams.
Nutshell
A comprehensive CRM integration allowing AI agents to manage B2B sales pipelines, contacts, and lead activities via the Membrane framework.
SuperOffice
A comprehensive CRM integration skill that allows AI agents to manage customer relationships, sales pipelines, and support tickets via the Membrane framework.
Workable
A powerful recruitment automation skill that connects your AI agent to the Workable applicant tracking system.
Qwen Cloud Model Selector
An advisory skill that recommends the optimal Qwen model and configuration based on task requirements, pricing, and performance goals.
Deployment Manager
A specialized AI agent skill designed to manage complex release orchestration, deployment strategies, and production reliability through automation.
Featured*
