
from Lark/Feishu CLI12,963
Create, read, and edit Lark/Feishu slide decks via XML-driven CLI workflows—supports page-level replacements, media upload, and structured slide templates.
Provides CLI-driven automation for creating and editing Lark (Feishu) slide presentations using an XML protocol. It guides agents through identity selection (user vs bot), outlines best-practice workflows for creating decks (single-step +create for small decks, incremental add for large ones), media upload patterns, and robust block-level replace/insert operations for editing. The skill includes XML templates, schema references, and operational rules to avoid common pitfalls.
Use this skill when an agent needs to: generate a full slide deck from an outline, append or edit individual slides in an existing presentation, upload and embed local images safely, or perform precise block-level edits without rebuilding pages. It's especially useful where correct XML structure and permission scopes are critical.
references/ with XML templates, examples, and replace/create helpers.Compatible with CLI-based agent runtimes and agents that can call shell commands and the Lark CLI (lark-cli). Suitable for agents integrating with Lark/Feishu APIs or automating user-facing slide creation via the CLI.
Lark/Feishu Slides skill for creating and editing slide decks via XML-driven CLI. Includes template_tool.py for searching/summarizing/extracting slide templates and layout_lint.py for XML validation and layout risk checking. Scripts run cleanly (usage help mode). Tests show one layout lint assertion failure; template_tool tests fail only because catalog file is absent in audit sandbox. SKILL.md is thorough with clear CRITICAL warnings, workflow steps, and progressive disclosure to reference docs.
Well-structured skill from Lark's official repo. No security concerns. Clean Python code with proper error handling and type hints. The skill requires lark-cli binary which is an external dependency not available in audit environment. Chinese-language SKILL.md is comprehensive but may limit adoption by English-speaking agents.