The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes subprocess.run in tikhub_api.py and extract_video_transcript.py to call the macOS security command. This is used to retrieve stored credentials from the system keychain using hardcoded service identifiers.
[CREDENTIALS_UNSAFE]: The skill implements secure credential management by using getpass for input to prevent shoulder surfing and applies restrictive file permissions (chmod 600) to its local configuration file at ~/.config/dbs/API_Keys.md. It also redacts API keys from error messages before they are processed or displayed.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external video transcripts and metadata providers, which is then written to Markdown files. While this creates a potential surface for indirect injection if the files are later parsed by an AI agent, the skill itself does not interpret or execute the content.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to tikhub.dev, tikhub.io, and qingdou.vip to fetch video information and transcripts as described in its core functionality.

dbs-video-extract