The Agent Skills Directory

[Prompt Injection] (SAFE): The content consists of instructional documentation and API implementation patterns. No override, bypass, or 'ignore previous instruction' markers were found.
[Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file paths (e.g., ~/.ssh, .env), or unauthorized network operations were identified.
[Obfuscation] (SAFE): No Base64 encoding, zero-width characters, or homoglyph-based evasion techniques were detected in the source text or code snippets.
[Unverifiable Dependencies & Remote Code Execution] (SAFE): The skill references the official anthropic Python library. No remote script execution (e.g., curl|bash) or dynamic code execution (eval/exec) patterns are present.
[Indirect Prompt Injection] (LOW):
Ingestion points: Patterns 1 and 3 involve processing large_document_content and retrieved_documents.
Boundary markers: None explicitly shown in the code examples.
Capability inventory: Frontmatter allows Bash, Write, and Task tools.
Sanitization: Not present in the snippets.
Assessment: While the skill describes processing external content while having access to high-privilege tools, it is purely a documentation guide and does not implement a vulnerable pipeline itself.

prompt-caching