Agent-first Marketing Image Generation

Generates and refines image prompts, estimates costs, and produces business-ready images with predictable pricing for agent workflows via Rynjer API.

This package implements the advertised image-generation flow and has a safe-looking mock mode, but there are important mismatches to review before enabling live mode: 1) The manifest declares no required env vars but the runtime/README require RYNJER_USE_LIVE, RYNJER_BASE_URL, and RYNJER_ACCESS_TOKEN for live operation — expect network calls to the default BASE_URL (https://rynjer.com) if you set live mode. 2) The IMPLEMENTATION.md describes an agent registration / owner-bind flow that issues a ryn_agent_v1_... key; understand and control who performs the owner bind so you don't inadvertently grant scopes. 3) Generation is paid: ensure any token you provide has minimal scopes and you understand pricing behavior (estimate vs generate). Recommendations before installing or running in live mode: - Start in mock mode (do not set RYNJER_USE_LIVE) to verify behavior locally. - If you enable live mode, only set RYNJER_ACCESS_TOKEN when you trust the Rynjer endpoint and have a token with minimal scopes. - Confirm the BASE_URL is the official service you expect; the code will use it directly. - Review the owner-bind/key creation process (out-of-band via UI) so you don't expose credentials to untrusted agents. - If you need stronger guarantees, ask the publisher to update registry metadata to declare the required env vars and clarify the exact auth steps and minimal scopes.

Type: OpenClaw Skill Name: rynjer-image-generation Version: 0.3.0 The rynjer-image-generation skill provides a professional and well-documented interface for image generation workflows, including prompt optimization, cost estimation, and result polling. The core logic in `src/mock-runtime.js` is clean and follows standard practices, using Node.js HTTPS requests to communicate with a dedicated backend (rynjer.com) while managing credentials via environment variables. No evidence of data exfiltration, malicious execution, or prompt injection was found in the code or the instructional markdown files.