The Agent Skills Directory

[COMMAND_EXECUTION]: The skill extracts file paths and dates from the ## Related Files table and header metadata in markdown documents, and then uses them as arguments in shell commands like git log and grep. If a document contains paths or dates with shell metacharacters, it could lead to the execution of arbitrary commands when the skill processes the file.\n- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection in its verification logic. It reads content from the ## Acceptance Criteria and ## Related Files sections of markdown files and interpolates it directly into a prompt for a sub-agent without sanitization or boundary markers.\n
Ingestion points: Acceptance Criteria and Related Files sections in docs/features/*/requests/*.md.\n
Boundary markers: Absent from the verification agent prompt template in SKILL.md.\n
Capability inventory: Read, Write, Bash, Agent (sub-agent execution), and Grep.\n
Sanitization: Absent; the skill does not escape or validate extracted text before interpolation into commands and prompts.

create-request