The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/_auth.py utilizes subprocess.run to execute a script from the sibling skill chanjing-credentials-guard. This mechanism is used to open a browser login page when local credentials are not found.
[DYNAMIC_EXECUTION]: Multiple scripts, including scripts/create_person.py, scripts/upload_file.py, and scripts/list_persons.py, utilize the __import__("os") function to dynamically load the os module for environment variable retrieval. This is used to determine API base URLs and credential paths.
[DATA_EXFILTRATION]: The skill facilitates the transmission of local video file content to dynamic sign_url endpoints provided by the API during the upload process. Additionally, the scripts/create_person.py script supports an optional --callback parameter which, if used, will cause the remote service to POST task results (containing sensitive resource references) to a user-specified URL.
[INDIRECT_PROMPT_INJECTION]: The skill processes JSON data retrieved from external API endpoints and interpolates it into script outputs.
Ingestion points: API response bodies in all execution scripts (e.g., scripts/get_person.py, scripts/poll_person.py).
Boundary markers: No specific boundary markers or instruction-ignore delimiters are present in the documentation or script prompts.
Capability inventory: The skill has network access (POST/PUT/GET), file read access (local videos, credentials), and file write access (updating local credential tokens).
Sanitization: The skill relies on standard json.loads parsing but does not implement additional sanitization or validation of the data returned by the API before it is processed or displayed.

chanjing-customised-person