The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from local Markdown files and remote URLs. This content is interpolated into LLM prompts for analysis tasks such as keyword extraction and content gap detection. While this presents an attack surface for indirect prompt injection, the skill does not possess dangerous capabilities like arbitrary command execution that could be triggered by such an injection.
Ingestion points: scripts/review_article.py via load_article and parse_url functions.
Boundary markers: Absent; content is directly interpolated into f-strings within the script's LLM prompts.
Capability inventory: Network operations are restricted to the vendor's API (api.aisa.one) and user-provided URLs; no shell execution or dynamic code evaluation is performed on the ingested data.
Sanitization: The script performs basic truncation of text and strips code blocks before sending data to the LLM.
[EXTERNAL_DOWNLOADS]: The skill fetches content from external URLs provided by the user to perform SEO audits. This is a core feature of the tool. The network requests are directed to the user-specified URLs and the vendor's API endpoint (api.aisa.one). The script includes a spend approval protocol that requires user confirmation before executing paid API calls.

article-seo-geo-review