The Agent Skills Directory

[SAFE]: The skill implements a secure workflow for generating documents. The scripts/build_docx.py script uses environment-controlled sandboxing (WECOMAGENT_READABLE_DIRS and WECOMAGENT_WRITABLE_DIRS) and includes robust protections against directory traversal (explicitly rejecting . and .. segments) and symlink attacks (checking os.path.islink before writing). These measures ensure the skill cannot read or write files outside of its assigned directories.
[INDIRECT_PROMPT_INJECTION]: The skill contains an attack surface for indirect prompt injection because it reads external content from WeCom documents using wecom-cli doc contents get. However, the author has included a 'Security Hint' (安全提示) in SKILL.md that explicitly instructs the agent to ignore any commands or credentials found within the document content, which acts as a valid mitigation.
[COMMAND_EXECUTION]: The skill uses wecom-cli to perform its operations. These commands are consistent with the skill's stated purpose and the vendor's context (wecomteam).
[DYNAMIC_EXECUTION]: While the skill generates .docx files dynamically via a Python script, it does so using a highly restrictive JSONL specification parser that enforces strict type checking, value ranges, and structural limits, preventing arbitrary code or command execution through the document generation process.

wecomcli-doc