The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses dynamic shell command substitution to resolve its plugin root and script paths at runtime. It employs the pattern ${CLAUDE_PLUGIN_ROOT:-$(ls -td "$HOME"/.claude/plugins/cache/insight-wave/cogni-trends/*/ | head -1)} to discover and execute local scripts. Assembling executable paths via shell subshells is a form of dynamic execution that can be exploited if the directory structure is manipulated.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its ingestion of external project data without proper sanitization or boundary markers.
Ingestion points: Data is ingested from the output of discover-projects.sh, project-status.sh, and project files like tips-project.json and trend-scout-output.json (SKILL.md).
Boundary markers: There are no explicit delimiters or safety instructions provided to the agent to distinguish between project data and instructions.
Capability inventory: The skill possesses extensive capabilities including Bash, Read, Write, Edit, Glob, and Grep (SKILL.md).
Sanitization: The instructions do not define any sanitization, escaping, or validation logic for the external content before it is processed or presented.

trends-resume