The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill demonstrates AI integration patterns that are susceptible to indirect prompt injection attacks.\n
Ingestion points: Untrusted data enters the context through the save_post hook (accessing $post->post_content) and the REST API input schema in references/detailed-guide.md.\n
Boundary markers: The code snippets for my_plugin_generate_ai_summary and my_plugin_generate_summary_cb interpolate the untrusted content directly into the prompt string using concatenation and sprintf without using delimiters or instructions to the model to ignore instructions within the data.\n
Capability inventory: The skill possesses the capability to perform AI text generation via wp_ai_client_prompt and write data to the WordPress database using update_post_meta.\n
Sanitization: Although standard WordPress sanitization functions like strip_tags and sanitize_textarea_field are used, they do not mitigate the risk of an LLM interpreting instructions hidden within the provided text.

wordpress-plugin-development