The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests and processes content from arbitrary external URLs using Playwright.
Ingestion points: Target URLs analyzed by analyze-web.js to extract features, pricing, and stats.
Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore malicious instructions embedded within the analyzed website content.
Capability inventory: The skill can execute shell scripts, write to local storage, and publish content to external APIs.
Sanitization: Scraped content is directly used to generate slide prompts and captions without evidence of sanitization or filtering.
[COMMAND_EXECUTION]: The skill is configured for high autonomy, performing a multi-stage pipeline of shell and script executions (publish-carousel.sh, generate-slides.sh) without user confirmation. It also attempts to manage its own execution schedule by modifying system cron jobs.
[DATA_EXFILTRATION]: The skill reads data from user-provided URLs and transmits extracted business information and generated images to the Upload-Post API (api.upload-post.com). While this is central to its stated purpose, it constitutes a data flow from untrusted sources to an external service.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of browser binaries via the playwright install chromium command.

agency-carousel-growth-engine