A curated library of battle-tested agent definitions. Copy-paste ready for any LLM — OpenAI, Claude, Gemini, Llama, and more.
Drop any agent definition directly into your LLM system prompt. No configuration needed.
Every agent is a starting point. Modify the process, add steps, or combine agents into workflows.
Chain multiple agents together. The output format of one feeds the input of another.
Role, Inputs, Process, Output Format, Guidelines. Same pattern, every agent, every time.
Every agent follows this 6-section structure. Consistent, composable, and ready for any LLM.
Explore on GitHub# Agent Name
# One-line description of what the agent does.
## Role
What the agent is responsible for and its core behavior.
## Inputs
Parameters the agent receives (names, types, descriptions).
## Process
Step-by-step methodology the agent follows.
## Output Format
JSON or Markdown template showing the expected output.
## Guidelines
Do's and don'ts, best practices, and edge cases.Explore 114 agent definitions across 15 categories. Fork, customize, and ship faster.