Bug Description

🐛 Issue Summary

When using Hermes Agent for CLI chat or gateway messaging (Telegram/Discord/Slack), the agent frequently throws the error:
Error: Response truncated due to output length limit
when generating long-form responses. This truncates the output mid-stream, breaks the conversation flow, and prevents users from receiving complete, usable answers.

✅ Expected Behavior

1. Long responses should be automatically split into chunks or streamed continuously to the user, rather than being hard-truncated with an error.
2. The system should handle output length limits gracefully (e.g., auto-continue the response, prompt the user to request more, or adjust context window dynamically) instead of interrupting the process.
3. The fix should work consistently across all use cases: regular chat, tool calls, code generation, and multi-step reasoning.

Steps to Reproduce

🔍 Minimal Reproduction Steps

1. Launch the Hermes Agent CLI with hermes chat (or start the gateway with hermes gateway start).
2. Send a prompt that requires a long, detailed response, for example:
   • "Please perform a full code review of this project, including architecture analysis, dependency issues, and optimization suggestions."
   • "Write a complete Python automated trading strategy with backtesting, risk management, and logging modules."
3. Observe the agent throw Error: Response truncated due to output length limit and cut off the response mid-generation.
4. Repeat the test with different long-form prompts to confirm the issue is consistent.

Expected Behavior

📌 Additional Context

• This error occurs reliably when the model generates long text, complex tool call chains, or full code implementations.
• I have already updated to the latest version via hermes update and confirmed the bug persists.
• I have searched existing issues and found no open report that fully addresses this specific error scenario.

Actual Behavior

🧪 Environment Details

• Hermes Agent version: [FILL IN YOUR VERSION, e.g., v0.4.2, run hermes --version to check]
• OS/Environment: [FILL IN, e.g., macOS 15.3, Ubuntu 22.04, Docker]
• LLM Provider/Model: [FILL IN, e.g., OpenRouter GPT-4o, local Ollama Llama 3 70B]
• Trigger Scenario: [FILL IN, e.g., CLI chat, Telegram gateway, Discord bot]

Affected Component

CLI (interactive chat)

Messaging Platform (if gateway-related)

No response

Operating System

macOS

Python Version

3.11.9

Hermes Version

2.1.0

Relevant Logs / Traceback

Root Cause Analysis (optional)

No response

Proposed Fix (optional)

No response

Are you willing to submit a PR for this?

• I'd like to fix this myself and submit a PR