The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to interact with a utility named wecom-cli using a command-line interface syntax (e.g., wecom-cli sheet create --json '...'). While the skill states this is a tool provided by the platform rather than a raw terminal command, the use of string-based command construction poses a risk if arguments are not properly escaped.
[DATA_EXFILTRATION]: The sheet import command in SKILL.md takes a file_path parameter. If an agent is coerced by a user to provide a path to sensitive system files (such as ~/.ssh/id_rsa or .env files), the agent will read these files and upload their contents to the Enterprise WeChat cloud service, effectively exfiltrating them.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface when handling external data.
Ingestion points: Data is ingested from local files during the import process and from remote spreadsheet content via wecom-cli sheet ranges get (as described in references/sheet-ranges-get.md).
Boundary markers: The instructions lack boundary markers or explicit guidance for the agent to ignore instructions embedded within the data retrieved from sheets or imported files.
Capability inventory: The skill possesses significant capabilities, including reading local files and creating or modifying cloud-based documents.
Sanitization: There is no evidence of content sanitization or validation to prevent malicious payloads within the spreadsheet data from influencing agent behavior.

wecomcli-sheet