The Agent Skills Directory

[SAFE]: The skill performs routine local filesystem operations including searching for the project root, reading PROGRESS.md, and creating backup files during format upgrades.
[COMMAND_EXECUTION]: The skill uses git to analyze commit history for automatic language detection (English/Chinese). This usage is restricted to local metadata and provides legitimate project-specific functionality.
[PROMPT_INJECTION]: The skill ingests untrusted content from the PROGRESS.md file. 1. Ingestion points: Reading PROGRESS.md from the project root. 2. Boundary markers: Absent; there are no explicit delimiters to isolate the file content from agent instructions. 3. Capability inventory: Filesystem write access for initializing the progress file or performing upgrades. 4. Sanitization: Absent; content is extracted based on headers and displayed directly. While this represents an indirect prompt injection surface, the risk is negligible because the logic is focused on parsing specific sections for display rather than executing content.

progress-show