feat: add Requesty Image Generation node

[Thibaultjaigu]

Summary

Adds a second node to the package: Requesty Image Generation, calling Requesty's OpenAI-compatible POST /v1/images/generations endpoint. The existing chat model node is unchanged in behavior.

• New ImageGenRequesty node: text prompt -> image, with model, size, quality, background, output format, and number-of-images options
• Binary output by default (preview in the n8n output panel, works with downstream binary nodes); Return Image URLs option for JSON output
• Usable as an AI Agent tool (usableAsTool) - the agent fills the prompt via $fromAI(); requires N8N_COMMUNITY_PACKAGES_ALLOW_TOOL_USAGE=true
• Uses httpRequestWithAuthentication with the existing requestyApi credential (no credential changes needed by users)
• Extracted shared attribution-header merging into nodes/shared/headers.ts, reused by both nodes
• Fix: credential authenticate() now only sets HTTP-Referer/X-Title when absent, so node-level custom headers can override them as documented
• Unit tests (33 passing), README docs, and a manual test workflow (test-workflow-image.json)
• Version bump to 1.1.0

Testing

• npm run lint, npm test, npm run build all clean
• Manually verified all three paths against the production gateway: binary generation, URL mode, and agent-tool usage (end-to-end run with gpt-5.4 calling the image tool)