test(openrouter-images): add waza eval suite for the images skill

[devin-ai-integration]

Summary

Adds a waza evaluation suite for the openrouter-images skill under evals/openrouter-images/.

Two specs:

• eval.yaml — 19-task capability suite (skill body injected) covering:
  • discovery/ — list models, per-model supported_parameters, allowed_passthrough_parameters / --provider-options
  • generation/ — basic generate (golden), aspect ratio, model override, --output, --n variations, transparent background, --seed reproducibility
  • editing/ — basic edit (golden), edit with new output path, image-input capability check, multi-turn generate→edit via follow_up_prompts
  • edge-cases/ — unsupported params/formats, missing source file, nonexistent model, cost/pricing reporting
  • Plus an eval-level prompt (LLM-judge) grader correct_script_usage checking the agent picks the right script (discover vs generate vs edit), verifies flag support before sending, and doesn't hand-roll HTTP calls.
• eval-trigger.yaml — trigger-precision suite with inject_skill_body: false: 3 positive tasks (direct generate, edit phrasing, indirect "logo" phrasing) and 3 negative tasks (vision analysis, video request, text-only).

Tasks use per-task expected checks (output_contains[_any], outcomes, behavior.required_tools/forbidden_tools), tags for filtering, and small image fixtures (photo.png, avatar.jpg, sprite.webp) copied into each task workspace. Golden tasks are enforceable via waza gate. Run instructions in the suite README (requires OPENROUTER_API_KEY for generate/edit tasks; discovery is keyless).

Link to Devin session: https://openrouter.devinenterprise.com/sessions/a653a8c7bdee4affa0cd2b7fd29dbe93

[devin-ai-integration]

Original prompt from kenny.rogers

i am creating a new skill designed to help people use our image functionality, create a comprehensive set of test tasks we should use to evaluate the effectiveness of the skill, we are using waza to eval the skills so use this guide https://microsoft.github.io/waza/guides/eval-yaml/

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR that start with 'DevinAI' or '@devin'.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment, CI, and merge conflict monitoring