The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The scripts/setup.sh file downloads an installer from https://astral.sh/uv/install.sh and pipes it directly to the shell to install the uv package manager.
[EXTERNAL_DOWNLOADS]: The skill clones the source code for the MediaCrawler tool from https://github.com/NanmiCoder/MediaCrawler.git and installs its dependencies.
[COMMAND_EXECUTION]: The manifest.json and scripts/setup.sh files execute several shell commands, including git clone, uv sync, uv run, and playwright install, to set up and manage the scraping environment.
[PROMPT_INJECTION]: The skill's primary function of collecting and displaying social media data introduces an indirect prompt injection surface (Category 8). Ingestion points: Untrusted social media content is collected from various platforms and stored in the data/ directory, later accessed via scripts/show_results.sh. Boundary markers: No delimiters or ignore-instructions are used to isolate external content from agent instructions. Capability inventory: The skill has the capability to execute shell commands and file system operations as defined in scripts/setup.sh and manifest.json. Sanitization: No evidence of data sanitization or validation is present before the collected content is presented to the agent.

media-crawler