> For the complete documentation index, see [llms.txt](https://underdogcowboy.gitbook.io/underdogcowboy-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://underdogcowboy.gitbook.io/underdogcowboy-docs/tools/agent-flow.md).

# Agent Flow

#### **Introduction**

**UnderdogCowboy** is a Python library designed to empower users with the flexibility to **configure, interact, and optimize LLMs (Large Language Models)** from various providers. At its core, the library enables knowledge workers to fine-tune AI interactions, aligning them with specific workflows and objectives without vendor lock-in. The library is complemented by a powerful TUI (Text User Interface) application, known as **Agent Flow**, which serves as an interactive workspace for managing, refining, and evaluating AI agents.

**Agent Flow** is divided into three specialized TUI screens, each tailored to a distinct aspect of agent development and interaction:

1. **Timeline Editor**: This screen facilitates seamless dialogs with LLMs, allowing users to craft conversations that can be saved locally either as **dialogs** or **agents**. It provides a robust environment for creating reusable interaction scripts, which can be reloaded and further refined.
2. **Agent Clarity**: Designed to sharpen the responses of AI agents, this tool helps guide the agent's outputs in a specific direction, ensuring that the responses align with the user's desired tone, context, and detail level. It's an essential feature for users looking to optimize the precision of their agents' responses.
3. **Agent Assessment Builder**: This screen allows users to construct a detailed **assessment structure**, which can later be used to evaluate the performance of agents in production. By defining custom assessment categories and criteria, users can automate the evaluation process, ensuring their agents consistently meet quality standards.

Together, these tools provide a comprehensive environment for users to **collaborate with AI**, iteratively refine agent behavior, and ensure optimal performance in real-world applications.

#### **Key Perspectives**

1. **Showcase for Building Custom Tools**:
   * The tools in Agent Flow are **demonstrations of what’s possible** using the UnderdogCowboy library. They illustrate how you can leverage the library to design tools tailored to your own unique workflows. By understanding how these tools work, you can **build similar applications** that align with your specific business processes, allowing you to harness the power of LLMs in ways that best fit your needs.
2. **Support for Manual Agent Development and Integration**:
   * These tools are designed to **assist you in the manual process of crafting and refining agents** before full integration into your automation scripts. Whether you're just starting out or refining an existing agent, the TUI screens help streamline the development process, making it easier to adjust and validate responses before committing to the underlying **Agent Dialog Manager**. This approach ensures that when you do integrate your agents into the automated workflows powered by the UnderdogCowboy library, they’re optimized for the best possible performance.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://underdogcowboy.gitbook.io/underdogcowboy-docs/tools/agent-flow.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.