The Agent Skills Directory

[PROMPT_INJECTION]: No malicious instructions, bypass markers, or attempts to override agent behavior were found. The skill instructions focus entirely on software quality assurance methodologies.
[CREDENTIALS_UNSAFE]: No hardcoded secrets or sensitive tokens are present. The skill provides clear guidance on using placeholders and implements patterns for filtering sensitive headers (e.g., authorization, x-api-key) from test artifacts to prevent accidental exposure.
[EXTERNAL_DOWNLOADS]: The skill references standard, industry-recognized testing tools and libraries such as Jest, Vitest, Playwright, and pytest. Installation commands provided are for official package registries and well-known services.
[DATA_EXFILTRATION]: No patterns for unauthorized data access or transmission were identified. Network operations mentioned are strictly for testing purposes using mocking frameworks like MSW and VCR.py to ensure deterministic behavior.
[INDIRECT_PROMPT_INJECTION]: The skill promotes secure data handling by including patterns for schema validation (Zod, Pydantic) and sanitization, which effectively reduces the surface for indirect injection attacks when the templates are used in practice.
[OBFUSCATION]: The skill does not contain any obfuscated content, encoded strings, or hidden text patterns designed to evade detection.

testing-strategy-builder