AI Agent Pattern

Also Known As

TL;DR

Explanation

An AI agent extends a simple LLM call by giving the model access to tools — functions it can invoke to take actions in the world. The agent receives a goal, reasons about what tool to call, receives the tool result, and repeats until it can produce a final answer. Common tools include web search, code execution, database queries, and API calls. The ReAct pattern (Reasoning + Acting) is the most common agent architecture: the model alternates between Thought (reasoning about what to do), Action (calling a tool), and Observation (receiving the result). In PHP, agents are typically built by calling an LLM API in a loop, parsing tool-call responses, executing the requested function, and feeding results back as the next message.

Common Misconception

Why It Matters

Common Mistakes

• No iteration limit — agents can loop indefinitely if the model repeatedly calls the wrong tool or misinterprets results.
• Giving the agent access to destructive tools without confirmation — file deletion, email sending, and database writes should require explicit user approval.
• Not logging tool calls — debugging an agent that produced a wrong answer requires replaying every reasoning step.
• Treating the agent as deterministic — the same input may take different tool-calling paths across runs, making testing difficult.

Avoid When

• Simple single-turn tasks where a standard LLM call is sufficient — agents add latency and cost per loop iteration.
• Irreversible actions without human confirmation — agents can take destructive actions based on misunderstood goals.
• Untrusted or ambiguous goals — an agent given a vague objective may pursue an unintended interpretation.
• Production environments without iteration limits and logging — runaway agents consume tokens and cause unintended side effects.

When To Use

• Multi-step tasks that require planning, tool use, and adapting based on intermediate results.
• Workflows where the path to the answer is not known upfront and the model must discover it.
• Automating research, code review, or data gathering tasks that a human would do step-by-step.
• When tool-use (web search, code execution, DB queries) is required to answer the goal.

Code Examples

// ❌ No iteration limit, no logging, destructive tool access
function runAgent(string $goal): string {
    while (true) {
        $response = callLLM($goal);
        if ($response['done']) return $response['answer'];
        $result = executeTool($response['tool'], $response['args']); // may delete files!
        $goal .= $result;
    }
    // infinite loop possible, no audit trail
}

<?php
// ✅ Agent with iteration limit, logging, and confirmation for destructive actions
function runAgent(string $goal, array $tools, int $maxIterations = 10): string
{
    $messages = [['role' => 'user', 'content' => $goal]];
    $iteration = 0;

    while ($iteration < $maxIterations) {
        $response = callLLM($messages, $tools);
        Log::info('Agent step', ['iteration' => $iteration, 'response' => $response]);

        if ($response['done']) return $response['answer'];

        // Require confirmation before destructive tools
        if (isDestructive($response['tool'])) {
            confirmWithUser($response['tool'], $response['args']);
        }

        $result = executeTool($response['tool'], $response['args']);
        $messages[] = ['role' => 'tool', 'content' => $result];
        $iteration++;
    }

    throw new RuntimeException("Agent exceeded max iterations ($maxIterations)");
}

References

• ↗ https://arxiv.org/abs/2210.03629

Tags

Related Terms