The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from project documents and user-provided feedback to generate retrospective reports.
Ingestion points: Reads project-specific files such as docs/03-增长迭代/迭代计划.md and docs/03-增长迭代/数据报告与用户反馈.md, and collects data via the AskUserQuestion tool.
Boundary markers: The skill lacks explicit boundary markers or instructions to ignore potential malicious prompts embedded within the processed documents or user feedback.
Capability inventory: The skill is granted Write permissions to the filesystem and Bash execution capabilities as defined in the allowed-tools section of SKILL.md.
Sanitization: No sanitization or validation logic is present to filter content from external documents before it is interpolated into the agent's context or used to generate the output report.

pm-retro