The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's execution flow constructs shell commands for the feishu-cli tool by interpolating parameters like <document_id>, <node_token>, and <output_path>. If these inputs are provided by a user or external source and are not properly sanitized or escaped, they could be leveraged for shell command injection.
[PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection attacks. It exports content from external Feishu/Lark documents and subsequently reads that content into the agent's context using the Read tool. Maliciously crafted content within those documents could attempt to override agent instructions. Ingestion points: Markdown files exported from Feishu to the local file system (e.g., in /tmp/). Boundary markers: No specific delimiters or safety instructions are defined to separate document content from system prompts. Capability inventory: The skill has access to the Bash and Read tools. Sanitization: No validation or sanitization of the document content is performed before the agent reads it.
[CREDENTIALS_UNSAFE]: The skill requires Feishu application credentials (FEISHU_APP_ID, FEISHU_APP_SECRET) which are stored in environment variables or a configuration file at ~/.feishu-cli/config.yaml. While no secrets are hardcoded in the skill itself, the reliance on and access to this sensitive configuration file by the underlying CLI tool represents a point of potential credential exposure.

feishu-cli-export