The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on shell commands including feishu-cli, jq, and python3 for its core functionality. It utilizes complex shell scripting patterns, such as loops and subshells, to automate multi-step document management tasks.
[DATA_EXPOSURE]: The skill interacts with the user's local authentication data stored in ~/.feishu-cli/token.json and requires management of Feishu API credentials via environment variables like FEISHU_APP_ID and FEISHU_APP_SECRET. This is a legitimate requirement for the tool's operation but involves the handling of sensitive configuration files.
[INDIRECT_PROMPT_INJECTION]: The skill presents an attack surface where malicious content within Feishu documents could potentially influence the agent's behavior during reading or summarization tasks.
Ingestion points: Content is retrieved from Feishu using doc export, wiki export, and sheet export commands as detailed in references/workflows/read/workflow.md.
Boundary markers: The instructions lack specific delimiters or instructions for the agent to ignore potentially malicious content embedded in the processed documents.
Capability inventory: The skill has access to the feishu-cli for cloud document operations, python3 for execution, and standard file Read/Write tools.
Sanitization: Document content is subjected to basic UTF-8 validity checks before processing.
[DYNAMIC_EXECUTION]: The skill uses python3 -c one-liners to perform runtime sanity checks on document encoding and content before executing import operations.

feishu-cli-docs