The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes data from an external third-party repository which creates an indirect prompt injection surface. Instructions embedded in the repository's README could influence the agent's behavior when it parses the file.\n
Ingestion points: The agent is instructed to clone and read the README.md file from github.com/pat-jj/Awesome-Adaptation-of-Agentic-AI.git.\n
Boundary markers: No boundary markers or instructions to ignore embedded commands are present in the parsing scripts.\n
Capability inventory: The skill utilizes subprocess.run, curl, and the requests library, providing a path for injected instructions to trigger network or system operations.\n
Sanitization: The provided Python and shell scripts do not sanitize the data extracted from the external README.md before processing it.\n- [EXTERNAL_DOWNLOADS]: The skill performs several network operations to external, non-vendor domains.\n
Clones a repository from https://github.com/pat-jj/Awesome-Adaptation-of-Agentic-AI.git.\n
Fetches metadata from the arXiv API at https://export.arxiv.org/api/query.\n
Includes a script (check_links.py) that performs HEAD requests to arbitrary URLs found within the untrusted repository.\n- [COMMAND_EXECUTION]: The skill provides numerous examples and scripts that execute shell commands and system calls.\n
Uses git clone, git pull, and git fetch to manage external code.\n
Employs shell utilities like grep, sed, and curl to manipulate and fetch data.\n
Includes Python scripts using subprocess.run to execute git commands and requests for network interaction.

awesome-adaptation-agentic-ai