Skip to content

fix: enable vision support for OpenAI-compatible client#153

Open
hazemahmedx0 wants to merge 3 commits into
mainfrom
fix/supports-vision
Open

fix: enable vision support for OpenAI-compatible client#153
hazemahmedx0 wants to merge 3 commits into
mainfrom
fix/supports-vision

Conversation

@hazemahmedx0
Copy link
Copy Markdown
Member

Summary

  • Sets supports_vision=True in the OpenAI-compatible client config (client.py:235)
  • Enables vision-capable models routed through this client to process image inputs

Test plan

  • Verify image inputs work end-to-end with an OpenAI-compatible vision model
  • Confirm non-vision model paths are unaffected

Set supports_vision=True so vision-capable models routed through
the OpenAI-compatible client can process image inputs.
@github-actions
Copy link
Copy Markdown

github-actions Bot commented May 11, 2026

All contributors have signed the CLA ✍️ ✅
Posted by the CLA Assistant Lite bot.

Adds a vision_format option ("openai" | "anthropic") to OpenAIProvider
and the message translation functions. Endpoints like MDB.AI that speak
Anthropic content format over an OpenAI-compatible HTTP envelope need
image blocks kept as-is rather than converted to image_url format.
@entelligence-ai-pr-reviews
Copy link
Copy Markdown

entelligence-ai-pr-reviews Bot commented May 11, 2026


Confidence Score: 5/5 - Safe to Merge

Safe to merge — this PR enables vision support for OpenAI-compatible clients, a well-scoped feature addition with no identified issues from automated or manual review analysis. The change appears clean with no logic bugs, security concerns, or regressions surfaced during review. No pre-existing unresolved comments are associated with this PR.

Key Findings:

  • No new review comments were generated, indicating the implementation passes all automated heuristic checks without any flagged concerns.
  • The PR has a narrow, well-defined scope — enabling vision support for an OpenAI-compatible client — which limits the blast radius of any potential undetected issues.
  • Zero critical, significant, or medium-severity issues were identified across all heuristic analysis dimensions, including logic, security, and correctness checks.

@hazemahmedx0
Copy link
Copy Markdown
Member Author

I have read the CLA Document and I hereby sign the CLA

github-actions Bot added a commit that referenced this pull request May 11, 2026
…d memory

When user_input is a list of content blocks (image + text), user_msg_str
was falling back to "" — losing the text context for system prompt
building and episodic memory logging. Extract the first text block instead.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant