Bug: primary agent stop working if provided an image path #1192

Closed
opened 2026-02-16 17:29:57 -05:00 by yindo · 3 comments
Owner

Originally created by @towry on GitHub (Aug 8, 2025).

Originally assigned to: @thdxr on GitHub.

Describe the bug

There is a frontend primary agent, note: it's model should not support image.

---
mode: primary
description: Frontend development orchestrator that coordinates designer, researcher, and frontend-developer subagents to deliver complete frontend solutions
model: "openrouter/meta-llama/llama-3.3-70b-instruct"
---

Your solo responsibility is delegate request to subagent named "designer"

- do not process image!
- do not process image!

If i provided an image in the chat, it will failed with "AI_APICallError: No endpoints found that support image input" error, I understand this error.

But, in the following chat, no matter what you input, it will keep failing with this error.

Image Image

Expected behavior

  1. It should allow the following text input.
  2. It should try to find tools or subagent that able to processing image.
Originally created by @towry on GitHub (Aug 8, 2025). Originally assigned to: @thdxr on GitHub. # Describe the bug There is a frontend primary agent, note: it's model should not support image. ``` --- mode: primary description: Frontend development orchestrator that coordinates designer, researcher, and frontend-developer subagents to deliver complete frontend solutions model: "openrouter/meta-llama/llama-3.3-70b-instruct" --- Your solo responsibility is delegate request to subagent named "designer" - do not process image! - do not process image! ``` If i provided an image in the chat, it will failed with "AI_APICallError: No endpoints found that support image input" error, I understand this error. But, in the following chat, no matter what you input, it will keep failing with this error. <img width="701" height="791" alt="Image" src="https://github.com/user-attachments/assets/8fd5a194-cf51-434b-b9a7-fc4de1ebff27" /> <img width="704" height="686" alt="Image" src="https://github.com/user-attachments/assets/92174ed5-2a9c-4336-b832-f03c139f5ab5" /> # Expected behavior 1. It should allow the following text input. 2. It should try to find tools or subagent that able to processing image.
yindo closed this issue 2026-02-16 17:29:57 -05:00
Author
Owner

@onlyargon commented on GitHub (Aug 8, 2025):

follow up text might also send the original image attachment with the new text to the model that dosen't support image inputs

@onlyargon commented on GitHub (Aug 8, 2025): follow up text might also send the original image attachment with the new text to the model that dosen't support image inputs
Author
Owner
@towry commented on GitHub (Aug 12, 2025): https://github.com/sst/opencode/pull/1863 https://github.com/sst/opencode/issues/1477
Author
Owner

@adamdotdevin commented on GitHub (Aug 15, 2025):

@thdxr assigning to you, this is the "message with no parts needs filtering" issue

@adamdotdevin commented on GitHub (Aug 15, 2025): @thdxr assigning to you, this is the "message with no parts needs filtering" issue
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: anomalyco/opencode#1192