The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/flake-detector.sh utility script uses unquoted variables ($GREP_FILTER, $PROJECT) within a shell command invocation of npx playwright. If an agent is manipulated into passing shell metacharacters through these arguments, it could lead to arbitrary command execution on the host machine.
[INDIRECT_PROMPT_INJECTION]: The skill provides scripts that serve as an attack surface for indirect prompt injection.
Ingestion points: Arguments passed to scripts/flake-detector.sh and scripts/generate-auth-profile.ts by the agent, likely derived from user-controlled test names or command-line parameters.
Boundary markers: None identified in the script execution flow to prevent argument manipulation.
Capability inventory: The agent can execute shell scripts and write to the file system via the provided utilities.
Sanitization: The scripts lack input validation or sanitization. Specifically, scripts/generate-auth-profile.ts allows for potential path traversal via the --role parameter used in file path construction, and scripts/flake-detector.sh allows for shell injection via unquoted variable interpolation.

playwright-e2e-testing