The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted user content, such as raw articles and external links, without utilizing boundary markers or sanitization. Malicious instructions embedded in the user input could potentially override the skill's defined SOP (Standard Operating Procedure).
Ingestion points: User-provided text and links in SKILL.md (Step 1).
Boundary markers: Absent; user content is directly interpolated into the generation process.
Capability inventory: File system write operations and external network requests.
Sanitization: Absent.
[EXTERNAL_DOWNLOADS]: The skill communicates with an external API to generate images based on user prompts.
Evidence: scripts/generate_image_seedream.py sends HTTP POST requests to https://api.gptnb.ai/v1/images/generations.
[COMMAND_EXECUTION]: The skill performs file system operations by creating and saving markdown files locally.
Evidence: SKILL.md (Step 4) instructs the agent to save the final blog post as a new .md file.

blog-post-writer