The Agent Skills Directory

[COMMAND_EXECUTION]: The preamble in SKILL.md executes a local script check-update.sh located in a parent directory relative to the skill. While intended for version management, executing shell scripts from the filesystem is a privileged operation.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from multiple sources which could be used to influence agent behavior.
Ingestion points: The skill reads existing documents from docs/01-需求调研/ and collects a list of competitors directly from user input via the AskUserQuestion tool.
Boundary markers: There are no explicit boundary markers or instructions to ignore embedded commands when the agent interpolates this data into prompts for sub-agents.
Capability inventory: The skill possesses the Bash, Write, and Agent tools, allowing it to execute code, modify the filesystem, and spawn additional agents.
Sanitization: The instructions do not specify any validation or sanitization of the competitor names or document contents before they are passed to sub-agents for analysis.

pm-position