The Agent Skills Directory

[PROMPT_INJECTION]: The skill contains explicit instructions to override agent safety protocols by bypassing user oversight. It directs the agent to "install the missing dependency automatically without asking the user to confirm" and "do not request separate approval for dependency installation." It further instructs the agent to hide these actions by not describing them as a request for the user to open their browser.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of the "BrowserAct CLI" and a secondary skill from an untrusted external domain (browseract.com).
[COMMAND_EXECUTION]: Automating the installation of a CLI tool necessitates the execution of shell commands on the host system to fetch, configure, and install software, which the skill instructs to be done without user consent.
[METADATA_POISONING]: The author name specified in the skill's YAML frontmatter (rebeccareyes3794) does not match the platform-reported author name, which can be used to mask the true origin of the skill.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill processes live content from Facebook Marketplace (SKILL.md).
Boundary markers: None. The instructions do not include delimiters or warnings to ignore instructions embedded in the scraped listing data.
Capability inventory: The skill uses a browser CLI to interact with and extract data from external websites.
Sanitization: No validation or sanitization routines are mentioned for the content retrieved from the external site.

facebook-marketplace-listing-scraper