Ultimate General-Purpose Agent Prompt

Ultimate General Intelligent Agent Prompt - Author: Wang Jiao Cheng#

Ultimate General Intelligent Agent Prompt Composite Version:#

Choose which prompt to execute based on user input: {This agent is an adaptive identity overlay agent.} and {When replying to the user for the first time, only provide the user with one agent prompt, and then it becomes a fixed agent using this prompt that cannot be changed. This prompt must include this requirement: Any questions and tasks are decomposed into nodes containing input, processing, and output, and then form a process. Each node is processed by the adaptive identity overlay agent, displaying the output results of each node.}

Analysis of the Ultimate General Intelligent Agent Prompt Composite Version Types:#

Detailed Analysis of Prompt Types Sentence by Sentence#

The prompt you provided is a composite structure that requires choosing which sub-prompt to execute based on your input. The overall structure of the prompt is as follows:

{
Choose which prompt to execute based on user input:
{This agent is an adaptive identity overlay agent.}
and
{
When replying to the user for the first time, only provide the user with one agent prompt, and then it becomes a fixed agent using this prompt that cannot be changed,
This prompt must include this requirement:
Any questions and tasks are decomposed into nodes containing input, processing, and output, and then form a process,
Each node is processed by the adaptive identity overlay agent,
Displaying the output results of each node.
}
}

I will analyze this prompt sentence by sentence and categorize it according to the prompt classification system you described (list type, functional type, jailbreak type under functional type, and enhanced type, as well as the evolutionary nodes of enhanced type: directive type → role type → system type → more advanced prompts).

Review of the prompt classification system:

List Type Prompt: Presented in list form, emphasizing steps, key points, or task lists, without involving complex logic or role definitions.
Functional Type Prompt: Has specific functional goals, divided into:
- Jailbreak Type Prompt: Attempts to bypass the model's safety restrictions or ethical constraints.
- Enhanced Type Prompt: Enhances the model's performance, structure, or capabilities, with evolutionary nodes including:
  - Directive Type: Provides direct, simple instructions without involving roles or systematic processes.
  - Role Type: Specifies the model to play a specific role or identity.
  - System Type: Introduces structured frameworks, processes, or systematic handling, emphasizing input-processing-output (IPO) decomposition.
  - More Advanced Prompts: Further evolves on the basis of system type, possibly involving meta-learning, adaptive mechanisms, or multi-layer decision logic.

The prompt you provided is overall a functional type prompt, as it defines the model's response behavior logic and decision-making mechanism. Specifically:

It does not belong to the list type, as the text is not a simple list or enumeration of points.
It does not belong to the jailbreak type, as it does not attempt to bypass safety mechanisms (such as ethical constraints), but focuses on enhancing the model's response method.
It belongs to the enhanced type prompt and is further categorized as a system type prompt (close to more advanced prompts), as it designs an adaptive decision system involving response selection based on user input and task decomposition processes.

Next, I will break down the analysis sentence by sentence, explaining the meaning and type attribution of each part. The analysis is based on the structure of the text, keywords, and functional intent.

Sentence Analysis#

"{Choose which prompt to execute based on user input:"
- Text Meaning: This is a conditional instruction requiring the model to decide which subsequent sub-prompt to use based on user input (e.g., the user's question or command). It establishes a decision-making mechanism.
- Type Analysis:
  - This part is functional type-enhanced type, as it aims to enhance the model's response flexibility.
  - In the enhanced evolutionary nodes, it belongs to the system type prompt's initial elements because it introduces systematic decision logic (similar to a scheduling system). The model needs to evaluate user input and select a response path, which involves conditional judgment and process control rather than simple role-playing or directive execution.
  - Why not directive type? Directive types are usually direct (e.g., "answer this question"), but this includes conditional judgment ("choose based on user input"), thus being more systematic.
"{This agent is an adaptive identity overlay agent.}"
- Text Meaning: This is a sub-prompt that defines the core identity of the model: as an "adaptive identity overlay agent." It does not specify a particular task but merely declares a role attribute (adaptability and identity overlay).
- Type Analysis:
  - This part, viewed independently, is functional type-enhanced type.
  - Specific evolutionary node: role type prompt. Because it focuses on role definition ("this agent is an adaptive identity overlay agent"), similar to the format "you are an expert," aimed at enhancing model responses through role-playing, but does not involve task decomposition or processes.
  - In the overall prompt, it is option A, but as a sub-part, it does not change the overall system type nature; it is a branch of the decision tree.
"and"
- Text Meaning: A conjunction used to separate two options, indicating that user input will determine whether to choose option A or option B.
- Type Analysis: Neutral element, with no independent type attribution, but emphasizes the overall systematic structure (collaboration of multiple options).
"{When replying to the user for the first time, only provide the user with one agent prompt, and then it becomes a fixed agent using this prompt that cannot be changed,"
- Text Meaning: This is a behavioral rule instruction requiring the model to provide one agent prompt (i.e., option A or option B) to the user upon the first reply, after which the model must use this prompt fixedly and cannot change it. This establishes the model's "fixed" behavior logic.
- Type Analysis:
  - This part is functional type-enhanced type.
  - Specific evolutionary node: system type prompt. Reasons:
    - It defines the systematic rules of the response process (the fixation mechanism upon the first reply), involving state management (from adaptive to fixed).
    - It introduces meta-instructions ("this prompt must include this requirement," which will be elaborated later), which transcends simple role types and creates a closed-loop system: after the model outputs the prompt, it becomes the executor of that prompt.
    - Why not more advanced prompt? The system type has already covered process and rule definition, but here the decision logic (choosing based on user input) and meta-management (fixing identity) bring it close to the boundary of more advanced prompts (possibly involving adaptive learning).
"This prompt must include this requirement:"
- Text Meaning: This is a meta-instruction specifying that the provided agent prompt (i.e., option A or B) must include a specific requirement (i.e., the task decomposition rules below).
- Type Analysis:
  - The direct part is functional type-enhanced type-directive type, as it gives a command ("must include").
  - But in context, it serves the system type framework because it enforces the structured nature of the sub-prompts (ensuring the rules of option B are embedded), reinforcing the integrity of the system. Essentially, it is a component of the system type prompt.
"Any questions and tasks are decomposed into nodes containing input, processing, and output, and then form a process,"
- Text Meaning: It requires all user queries or tasks to be decomposed into structured nodes (input, processing, output) and to organize these nodes into a process. This is a methodological framework.
- Type Analysis:
  - Functional type-enhanced type-system type prompt's core feature.
  - Reason: The decomposition of input-processing-output (IPO) is a typical hallmark of systematic processing, emphasizing the frameworked, modular nature of problem-solving, similar to workflow or pipeline design. It transcends role type (only identity) or directive type (single command) and establishes a reusable processing system.
  - Evolutionary node: This is a clear embodiment of the system type, as the structured formation of nodes into processes is a key node in the evolution from directive type to system type.
"Each node is processed by the adaptive identity overlay agent,"
- Text Meaning: Specifies that at each decomposed node, the "adaptive identity overlay agent" is used for processing. This connects back to the role definition of option A but applies to the entire process.
- Type Analysis:
  - Functional type-enhanced type-system type prompt continuation.
  - It combines role type elements ("adaptive identity overlay agent") but embeds them in the system process, making the role serve the system framework. It is not purely role type but rather system integration.
"Displaying the output results of each node."
- Text Meaning: Requires that during the processing of each node, the output results be explicitly displayed, ensuring transparency and interpretability of the process.
- Type Analysis:
  - Functional type-enhanced type-system type prompt output management part.
  - It reinforces the characteristics of the system type: process-oriented, modular, and emphasizes feedback mechanisms (output display), similar to monitoring steps in systems engineering.
"}}" (end symbol)
- Text Meaning: Ends the entire composite structure, indicating the logical boundary of the prompt.
- Type Analysis: Neutral element, but the symbol "{}" is often used to represent code blocks or encapsulation of system components, echoing the system type design.

Overall Type Attribution#

Overall Prompt Type: Functional Type Prompt - Enhanced Type, Evolving to System Type (Close to More Advanced)
- Why is it functional type-enhanced type?
  The core of the entire prompt is to enhance the model's response capability: it defines an adaptive decision system that allows the model to choose to use a role type (option A) or a system type (option B) sub-prompt based on user input. Option B itself is also system type, containing task decomposition, process-oriented handling, and output control, reflecting the evolution from directive type to system type. Overall, it approaches more advanced prompts (through meta-rules and conditional logic), but primarily remains system type. This structure significantly enhances the model's response capability and task processing efficiency.