coding-opencode

What This Skill Does

The coding-opencode skill acts as a bridge between the OpenClaw agent and the "Oh My OpenCode" ecosystem. It provides an advanced, agentic interface for software engineering tasks. Rather than relying on simple text generation, this skill leverages a multi-agent orchestration architecture—including specialized agents like Sisyphus for management, Hephaestus for deep work, and Oracle for technical design—to handle complex development workflows. It integrates directly with LSP (Language Server Protocol) and AST (Abstract Syntax Tree) tools, ensuring that code refactoring and generation are syntactically accurate and context-aware. Whether you are performing a simple script generation or orchestrating a large-scale codebase refactor, this skill manages the lifecycle of the task from start to finish.

Installation

To integrate this capability into your OpenClaw environment, execute the following command in your terminal:

clawhub install openclaw/skills/skills/iqbalnaveliano/coding-opencode

Ensure your local environment has the necessary permissions to interface with your filesystem, as this skill interacts heavily with code repositories to perform analysis and modification.

Use Cases

This skill is best suited for:

• Full-Stack Development: Generating boilerplate, logic, and UI components using multi-model orchestration.
• Refactoring & Migration: Using AST-based analysis to safely modify legacy codebases without breaking dependencies.
• Deep Debugging: Allowing the Oracle and Hephaestus agents to investigate complex runtime errors by traversing the codebase.
• Automated Documentation: Utilizing the Librarian agent to scan modules and generate or update documentation based on existing implementations.
• Autonomous Projects: Activating 'ultrawork' mode for end-to-end task completion when you need the AI to handle planning, implementation, and verification steps autonomously.

Example Prompts

1. "opencode --agent build --ultrawork 'Create a FastAPI boilerplate project structure with SQLAlchemy integration for PostgreSQL.'"
2. "opencode --agent build 'ulw: Audit the src/auth.py file for potential security vulnerabilities and suggest a refactor to improve performance.'"
3. "opencode --agent build 'Librarian, find all occurrences of legacy fetch calls in the frontend directory and outline the steps to migrate them to Axios.'"

Tips & Limitations

• Safety First: Always review code changes generated by 'ultrawork' modes. While the agentic approach is highly effective, large-scale automated refactoring should be tracked via Git version control.
• Performance: 'ultrawork' can be resource-intensive. Ensure you are not running it on critical production files without proper testing cycles.
• Context Windows: For extremely large repositories, use the Librarian or Explorer tools first to narrow down the scope before running complex refactoring agents.
• Consistency: If you find the agent struggling with logic, try explicitly naming the agent (e.g., calling Hephaestus) rather than relying on the general 'build' command for better specialization.