The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions focus on Turborepo task orchestration and configuration best practices. No attempts to override agent behavior, bypass safety filters, or extract system prompts were identified.
[DATA_EXPOSURE]: Technical references mention common environment variable names (e.g., TURBO_TOKEN, DATABASE_URL, API_KEY) within an educational context to illustrate how Turborepo handles hashing and secrets. No hardcoded credentials, sensitive personal information, or unauthorized file access patterns were found.
[REMOTE_CODE_EXECUTION]: The skill suggests standard development commands, such as npx turbo-ignore, and package installations via official registries (npm, pnpm, yarn, bun). No suspicious piped execution patterns (e.g., curl | bash) or remote script executions from unknown sources were identified.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and assist with project-specific files such as turbo.json and package.json. While this ingestion of local repository data is a standard surface for indirect injection, the skill maintains clear boundaries using markdown structure and focuses on configuration analysis rather than executing arbitrary instructions found in data.
[OBFUSCATION]: Analysis of the skill body and reference files found no hidden content, Base64-encoded commands, zero-width characters, or homoglyph-based URL spoofing.
[COMMAND_EXECUTION]: The skill recommends standard shell commands for monorepo management (e.g., turbo run, turbo watch). These commands are essential to the skill's primary purpose and do not involve privilege escalation or persistence mechanisms.

turborepo