The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the lark-cli command-line tool to perform document operations. It executes commands such as lark-cli docs +fetch, lark-cli docs +update, and lark-cli drive +add-comment to interact with the Feishu API. These operations are essential for the skill's primary function of reviewing and annotating documents.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface as it processes untrusted data from external Feishu documents and has write capabilities back to the platform.
Ingestion points: Document content is ingested via lark-cli docs +fetch --doc "<文档URL或ID>" in SKILL.md.
Boundary markers: The instructions do not specify explicit delimiters or warnings to ignore instructions embedded within the paper being reviewed.
Capability inventory: The skill has the capability to modify document content via lark-cli docs +update and post comments via lark-cli drive +add-comment as detailed in SKILL.md and references/revision-marks.md.
Sanitization: There is no mention of sanitizing or filtering the ingested document content before processing.
[SAFE]: The identified risks are inherent to the skill's primary purpose of document review and editing. No malicious patterns, such as data exfiltration to unknown domains, obfuscation, or unauthorized persistence mechanisms, were detected.

feishu-paper-reviewer