llama cpp error with superpower

[feyleth]

Description

issue on superpower

when launch with opencode and llamacpp i got error for parsing Jinja template

Llama cpp Logs

[33553] srv    operator(): got exception: {"error":{"code":400,"message":"Unable to generate parser for this template. Automatic parser generation failed: \n------------\nWhile executing CallExpression at line 85, column 32 in source:\n...first %}↵            {{- raise_exception('System message must be at the beginnin...\n                                           ^\nError: Jinja Exception: System message must be at the beginning.","type":"invalid_request_error"}}
[33553] srv  log_server_r: done request: POST /v1/chat/completions 127.0.0.1 400
srv  log_server_r: done request: POST /v1/chat/completions 127.0.0.1 400
[33553] srv    operator(): got exception: {"error":{"code":400,"message":"Unable to generate parser for this template. Automatic parser generation failed: \n------------\nWhile executing CallExpression at line 85, column 32 in source:\n...first %}↵            {{- raise_exception('System message must be at the beginnin...\n                                           ^\nError: Jinja Exception: System message must be at the beginning.","type":"invalid_request_error"}}

model
use Qwen3.5-35B-A3B

Plugins

superpower

OpenCode version

1.2.26

Steps to reproduce

1. install superpowers on opencode
2. try to chat with bot

Screenshot and/or share link

No response

Operating System

linux

Terminal

No response

[github-actions]

This issue might be a duplicate of existing issues. Please check:

• Multiple system prompts break Qwen3.5-* models #15059: Multiple system prompts break Qwen3.5-* models — same root cause: a plugin adding an extra system prompt triggers the Jinja System message must be at the beginning error with Qwen3.5 models
• When using the qwen3.5-122b-a10b model (via nvidia provider), the following error is returned: System message must be at the beginning #16560: Same error message System message must be at the beginning with a Qwen3.5 model via a different provider
• OpenCode sends tools (and Jinja tool template) to llama.cpp results in 500 error unless --jinja; with --jinja, template crashes (reject filter) #1890: llama.cpp + Jinja template errors when using OpenAI-compatible providers

If your issue is not resolved by checking those, feel free to keep this open and provide additional context on how it differs.

[github-actions]

To stay organized issues are automatically closed after 60 days of no activity. If the issue is still relevant please open a new one.