a0-development

Guides developers in extending the BioDockify AI framework with tools, extensions, and plugins.

2|Updated May 12, 2026
One-click install
npx skills add https://github.com/tajo9128/BioDockify-Pharma-AI --skill a0-development-tajo9128
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: a0-development
Source: https://github.com/tajo9128/BioDockify-Pharma-AI/tree/main/skills/a0-development
Command: npx skills add https://github.com/tajo9128/BioDockify-Pharma-AI --skill a0-development-tajo9128

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Developers extending the BioDockify AI framework often struggle to find correct paths, imports, base classes, and hook points, leading to broken tools, misplaced extensions, and incompatible plugins. ## Core Features & Use Cases - Framework Architecture Reference: Documents the full project layout, agent loop, plugin-first design, and user-space separation so code lands in the right directories. - Component Creation Guides: Provides working code patterns for Tools (Tool/Response base classes), Extensions (30+ lifecycle hook points with numeric ordering), Flask API endpoints, agent profiles, and prompt fragments. - Use Case: A developer wants to add a new agent capability. The skill shows them to create a Tool subclass in a plugin under /a0/usr/plugins/, add a matching agent.system.tool.<name>.md prompt fragment, and follow async execution conventions. ## Quick Start Ask the agent to load the a0-development skill and explain how to create a new tool with its prompt fragment for the BioDockify AI framework.

Frequently Asked Questions about a0-development

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create a custom tool for the BioDockify AI agent framework?

Create a class inheriting from Tool in helpers.tool, implement the async execute() method returning a Response, and place it in a plugin's tools/ directory. You must also add a matching agent.system.tool.<name>.md prompt fragment so the agent knows how to use it.

How do I add an extension to hook into the agent lifecycle?

Create a Python file named _NN_name.py under extensions/python/<hook_point>/, subclassing Extension from helpers.extension. The numeric prefix controls execution order, and hook points include agent_init, before_main_llm_call, tool_execute_before, and monologue_end.

What is the difference between plugins and extensions in BioDockify AI?

Plugins are the primary extension mechanism, bundling tools, extensions, prompts, and API endpoints into one package. Extensions are individual lifecycle hooks that can also live standalone, but the framework recommends packaging new capabilities as plugins.

Where should I put custom code so it survives framework updates?

Place user code under /a0/usr/: plugins in usr/plugins/, agent profiles in usr/agents/, skills in usr/skills/, and extensions in usr/extensions/. Everything under /a0/usr/ is user-space and survives framework updates.

How does the prompt override system work for agent profiles?

Profiles inherit prompts from the default profile, and placing a file with the same name in the profile's prompts/ directory overrides it. The framework searches profile-specific directories first, then plugins, then core prompts, using the first match found.