a0-create-agent

Create BioDockify AI agent profiles with agent.yaml, prompt overrides, tools, and extensions.

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Creating a new subordinate agent profile in the BioDockify AI (Agent Zero) framework requires knowing where profiles live, the exact agent.yaml schema, and how prompt inheritance works. This Skill guides the process end-to-end so profiles are valid, discoverable, and correctly scoped. ## Core Features & Use Cases - Progressive intake interview: Asks only the highest-value questions first, inferring sensible defaults instead of overwhelming the user with a long form. - Blueprint-first generation: Produces a strict AgentProfileBlueprint JSON (schema agent_zero.agent_profile_blueprint.v1) as the single source of truth before writing any files. - Full customization surfaces: Covers agent.yaml metadata, agent.system.main.specifics.md prompt overrides, profile-scoped _model_config LLM settings, custom Python tools, and lifecycle extensions. - Use Case: A user says "create a data analyst agent". The Skill interviews briefly, confirms a blueprint, then writes /a0/usr/agents/data-analyst/agent.yaml plus a tailored specifics.md prompt, ready for delegation via call_subordinate. ## Quick Start Ask the assistant to create a new agent profile for your desired specialty and confirm the proposed blueprint before files are written.

Frequently Asked Questions about a0-create-agent

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

FAQPage Schema
How do I create a new agent profile in Agent Zero?

Create a directory under /a0/usr/agents/<name> containing an agent.yaml file with title, description, and context fields. Add a prompts/agent.system.main.specifics.md file to define the agent's role and behavior, then delegate to it via call_subordinate using the directory name.

What fields does agent.yaml support in Agent Zero?

agent.yaml supports exactly three fields: title, description, and context. Model settings, temperature, and allowed_tools must not be added there; profile-specific models live in a companion _model_config/config.json file instead.

Where should user agent profiles be placed?

User profiles belong in /a0/usr/agents/<profile_name>/. The /a0/agents/ directory is reserved for core framework profiles like default, developer, and researcher, while plugin-distributed profiles go under /a0/usr/plugins/<plugin>/agents/.

Can an agent profile use a different LLM than the global default?

Yes, by creating a profile-scoped plugins/_model_config/config.json inside the profile directory. The file must contain complete chat_model, utility_model, and embedding_model sections because scoped configs are not deep-merged with global settings.

Why is my new agent profile not showing up?

Check that the directory name uses only lowercase letters, numbers, hyphens, or underscores and is unique across search paths. Also verify agent.yaml parses as valid YAML and spawn a fresh agent, since profiles load at agent initialization.

How do I add a custom tool to one agent profile only?

Place a Python Tool subclass in the profile's tools/ directory and add a matching prompts/agent.system.tool.<name>.md file describing its usage and JSON call schema. A file named like a core tool replaces that core tool for this profile only.