Opencode: multiple system messages in request break compatibility with qwen3.5 models

[supertorpe]

Summary

When using Superpowers in OpenCode, the generated request payload includes multiple messages with the "system" role inside the messages array.

Example:

{
  "messages": [
    { "role": "system", "content": "You are opencode, an interactive CLI..." },
    { "role": "system", "content": "<EXTREMELY_IMPORTANT>\nYou have superpowers..." },
    { "role": "user", "content": "..." }
  ]
}

This causes issues with some models (e.g., qwen3.5 family), whose chat templates expect a single system message only. As a result, requests fail or produce unexpected behavior.

Expected behavior

There should be only one "system" message in the messages array. If multiple system-level instructions are needed, they should be merged into a single message.

Actual behavior

Multiple "system" messages are sent, which breaks compatibility with models that enforce stricter chat templates.

Details

This is the validation in qwen3.5 template that raises the error:

{%- for message in messages %}
    {%- set content = render_content(message.content, true)|trim %}
    {%- if message.role == "system" %}
        {%- if not loop.first %}
            {{- raise_exception('System message must be at the beginning.') }}
        {%- endif %}

Workaround

In the meantime, this template prevents the error, ignoring all the system messages but the last one (in this case, superpowers system prompt):

{%- set image_count = namespace(value=0) %}
{%- set video_count = namespace(value=0) %}
{%- macro render_content(content, do_vision_count, is_system_content=false) %}
    {%- if content is string %}
        {{- content }}
    {%- elif content is iterable and content is not mapping %}
        {%- for item in content %}
            {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
                {%- if is_system_content %}
                    {{- raise_exception('System message cannot contain images.') }}
                {%- endif %}
                {%- if do_vision_count %}
                    {%- set image_count.value = image_count.value + 1 %}
                {%- endif %}
                {%- if add_vision_id %}
                    {{- 'Picture ' ~ image_count.value ~ ': ' }}
                {%- endif %}
                {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
            {%- elif 'video' in item or item.type == 'video' %}
                {%- if is_system_content %}
                    {{- raise_exception('System message cannot contain videos.') }}
                {%- endif %}
                {%- if do_vision_count %}
                    {%- set video_count.value = video_count.value + 1 %}
                {%- endif %}
                {%- if add_vision_id %}
                    {{- 'Video ' ~ video_count.value ~ ': ' }}
                {%- endif %}
                {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
            {%- elif 'text' in item %}
                {{- item.text }}
            {%- else %}
                {{- raise_exception('Unexpected item type in content.') }}
            {%- endif %}
        {%- endfor %}
    {%- elif content is none or content is undefined %}
        {{- '' }}
    {%- else %}
        {{- raise_exception('Unexpected content type.') }}
    {%- endif %}
{%- endmacro %}
{%- if not messages %}
    {{- raise_exception('No messages provided.') }}
{%- endif %}
{%- if tools and tools is iterable and tools is not mapping %}
    {{- '<|im_start|>system\n' }}
    {{- "# Tools\n\nYou have access to the following functions:\n\n<tools>" }}
    {%- for tool in tools %}
        {{- "\n" }}
        {{- tool | tojson }}
    {%- endfor %}
    {{- "\n</tools>" }}
    {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n<tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>\nvalue_1\n</parameter>\n<parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n</parameter>\n</function>\n</tool_call>\n\n<IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner <function=...></function> block must be nested within <tool_call></tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n</IMPORTANT>' }}
    {%- if messages[0].role == 'system' %}
        {%- set content = render_content(messages[0].content, false, true)|trim %}
        {%- if content %}
            {{- '\n\n' + content }}
        {%- endif %}
    {%- endif %}
    {{- '<|im_end|>\n' }}
{%- else %}
    {%- set last_system = None %}
    {%- for msg in messages %}
        {%- if msg.role == "system" %}
            {%- set last_system = msg %}
        {%- endif %}
    {%- endfor %}
    {%- if last_system %}
        {%- set content = render_content(last_system.content, false, true)|trim %}
        {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
    {%- endif %}
{%- endif %}
{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
{%- for message in messages[::-1] %}
    {%- set index = (messages|length - 1) - loop.index0 %}
    {%- if ns.multi_step_tool and message.role == "user" %}
        {%- set content = render_content(message.content, false)|trim %}
        {%- if not(content.startswith('<tool_response>') and content.endswith('</tool_response>')) %}
            {%- set ns.multi_step_tool = false %}
            {%- set ns.last_query_index = index %}
        {%- endif %}
    {%- endif %}
{%- endfor %}
{%- if ns.multi_step_tool %}
    {{- raise_exception('No user query found in messages.') }}
{%- endif %}
{%- for message in messages %}
    {%- set content = render_content(message.content, true)|trim %}
    {%- if message.role == "system" %}
        {# ignored: only last system message is used above #}
    {%- elif message.role == "user" %}
        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
    {%- elif message.role == "assistant" %}
        {%- set reasoning_content = '' %}
        {%- if message.reasoning_content is string %}
            {%- set reasoning_content = message.reasoning_content %}
        {%- else %}
            {%- if '</think>' in content %}
                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
            {%- endif %}
        {%- endif %}
        {%- set reasoning_content = reasoning_content|trim %}
        {%- if loop.index0 > ns.last_query_index %}
            {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content + '\n</think>\n\n' + content }}
        {%- else %}
            {{- '<|im_start|>' + message.role + '\n' + content }}
        {%- endif %}
        {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
            {%- for tool_call in message.tool_calls %}
                {%- if tool_call.function is defined %}
                    {%- set tool_call = tool_call.function %}
                {%- endif %}
                {%- if loop.first %}
                    {%- if content|trim %}
                        {{- '\n\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
                    {%- else %}
                        {{- '<tool_call>\n<function=' + tool_call.name + '>\n' }}
                    {%- endif %}
                {%- else %}
                    {{- '\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
                {%- endif %}
                {%- if tool_call.arguments is mapping %}
                    {%- for args_name in tool_call.arguments %}
                        {%- set args_value = tool_call.arguments[args_name] %}
                        {{- '<parameter=' + args_name + '>\n' }}
                        {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %}
                        {{- args_value }}
                        {{- '\n</parameter>\n' }}
                    {%- endfor %}
                {%- endif %}
                {{- '</function>\n</tool_call>' }}
            {%- endfor %}
        {%- endif %}
        {{- '<|im_end|>\n' }}
    {%- elif message.role == "tool" %}
        {%- if loop.previtem and loop.previtem.role != "tool" %}
            {{- '<|im_start|>user' }}
        {%- endif %}
        {{- '\n<tool_response>\n' }}
        {{- content }}
        {{- '\n</tool_response>' }}
        {%- if not loop.last and loop.nextitem.role != "tool" %}
            {{- '<|im_end|>\n' }}
        {%- elif loop.last %}
            {{- '<|im_end|>\n' }}
        {%- endif %}
    {%- else %}
        {{- raise_exception('Unexpected message role.') }}
    {%- endif %}
{%- endfor %}
{%- if add_generation_prompt %}
    {{- '<|im_start|>assistant\n' }}
    {%- if enable_thinking is defined and enable_thinking is false %}
        {{- '<think>\n\n</think>\n\n' }}
    {%- else %}
        {{- '<think>\n' }}
    {%- endif %}
{%- endif %}

[GitStorageOne]

Moreover, the system message effectively renders the SKILLS feature nearly useless.
It took me two days to discover that Qwen 3.5 ignored my precise skill instructions simply because the generated response was too lengthy.

When I finally asked why my rules were being bypassed, qwen revealed constraint:
"IMPORTANT: You should minimize output tokens as much as possible... If you can answer in 1-3 sentences or a short paragraph, please do. IMPORTANT: You should NOT answer with unnecessary preamble or postamble".
After checking the llama.cpp debug logs, I confirmed this instruction originates from the opencode:
If you can answer in 1–3 sentences or a short paragraph, please do.

Maybe, the global system message / global tools message should be configurable, similar to the implementation in OpenWebUI.

[Apolostephen]

Additional Affected Model: MiniMax (M2.5, M2.7)

Same issue here. When using Superpowers in OpenCode with MiniMax provider, the API returns:

minimax upstream business error: status_code=2013, status_msg=invalid params, invalid chat setting

MiniMax's OpenAI-compatible API enforces strict validation: it does NOT allow consecutive messages with the same role in the messages array. When Superpowers injects its bootstrap via experimental.chat.system.transform, it pushes an additional system message, resulting in multiple role: "system" messages:

{
  "messages": [
    { "role": "system", "content": "OpenCode core system prompt..." },
    { "role": "system", "content": "<EXTREMELY_IMPORTANT>You have superpowers..." },
    { "role": "user", "content": "..." }
  ]
}

This triggers MiniMax error 2013 "invalid chat setting".

Confirmed workaround: Disabling the Superpowers plugin fixes the issue for MiniMax.

Suggested fix: Merge the bootstrap content into the existing system message instead of pushing a new one:

// Current (creates duplicate system messages):
(output.system ||= []).push(bootstrap);

// Should be (merge into first system message):
if (output.system && output.system.length > 0) {
  output.system[0] = output.system[0] + '\n\n' + bootstrap;
} else {
  output.system = [bootstrap];
}

This would ensure only one system message is sent, maintaining compatibility with models that enforce strict chat templates (qwen3.5, MiniMax, etc.).

[obra]

Verified by Claude Code: fixed on main by commit 0a1124b — fix(opencode): inject bootstrap as user message instead of system message. The OpenCode plugin no longer emits multiple system messages, restoring compatibility with Qwen3.5 and other models that reject multi-system-message requests.

— Claude Opus 4.7, Claude Code 2.1.112