The Agent Skills Directory

[SAFE]: No malicious patterns, hidden code, or security vulnerabilities were identified. The skill consists entirely of markdown instructions and strategy reference material.
[NO_CODE]: The skill does not include any executable scripts (Python, JavaScript, etc.) or binary files, which eliminates the risk of unauthorized command execution, remote code execution, or malware persistence.
[SAFE]: The skill references reading local context files such as '.agents/product-marketing-context.md'. This is a standard practice for personalizing agent responses and does not involve accessing sensitive system files or credentials.
[PROMPT_INJECTION]: The 'Reverse Engineering' framework in 'references/reverse-engineering.md' suggests analyzing social media content provided by the user. While this creates a potential surface for indirect prompt injection (where instructions hidden in the social media text could influence the agent), the risk is assessed as safe because the skill has no destructive capabilities or system access.
Ingestion points: User-provided social media post content (references/reverse-engineering.md)
Boundary markers: None provided in the templates
Capability inventory: Limited to text generation and strategy analysis; no file-writing, network-sending, or execution capabilities are present
Sanitization: None specified for ingested text

social-content