The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from multiple sources, which could contain instructions designed to influence the agent's behavior.
Ingestion points: The skill parses user-provided input from $ARGUMENTS in SKILL.md (Step 1) and reads external project source code (components, configs, fixtures) using a subagent in SKILL.md (Step 2).
Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following malicious instructions embedded within the source code files it explores.
Capability inventory: The skill has the capability to execute shell commands via npx playwright test as seen in SKILL.md (Step 7).
Sanitization: The skill does not implement any validation or sanitization of the context gathered from the codebase before interpolating it into the generated test code.
[DYNAMIC_EXECUTION]: The skill implements a workflow that involves generating executable TypeScript or JavaScript files from templates and then running them. Generating code based on potentially untrusted project context and user input for immediate execution is a dynamic execution pattern.
[COMMAND_EXECUTION]: In SKILL.md (Step 7), the skill explicitly instructs the agent to run the generated test files using the command line (npx playwright test <generated-file>).

generate