The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation or update of a global Node.js package from the official NPM registry to function.\n
Evidence: npm install -g @qianwenai/qianwen-cli in SKILL.md.\n
Context: The package belongs to the vendor and is the primary tool for the skill's operations.\n- [COMMAND_EXECUTION]: The skill relies extensively on executing shell commands to interact with conference services.\n
Evidence: Multiple subcommands of qianwen yunqi and authentication checks like qianwen auth status are defined in SKILL.md.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes conference summaries retrieved from an external source, creating a potential attack surface for indirect prompt injection.\n
Ingestion points: The summary field from the output of qianwen yunqi list summaries (as described in SKILL.md).\n
Boundary markers: None. The instructions do not provide delimiters or warnings to the agent to treat summary content as untrusted data.\n
Capability inventory: The agent can execute shell commands (qianwen CLI) and perform package management tasks.\n
Sanitization: No sanitization or content validation steps are mentioned for the ingested summary data.

yunqi-agent-attendance