The Agent Skills Directory

[SAFE]: The skill provides strategic frameworks and educational content related to SaaS pricing. A thorough analysis of all instructions, evaluation files, and reference documents revealed no patterns associated with prompt injection, obfuscation, or remote code execution.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from local marketing context files and user prompts. However, it lacks high-risk capabilities, which significantly minimizes potential impact.
Ingestion points: Reads .agents/product-marketing.md, .claude/product-marketing.md, or product-marketing-context.md to gather business context (SKILL.md).
Boundary markers: No specific delimiters or instructions are used to distinguish external data from core instructions.
Capability inventory: The skill's operations are limited to generating text-based advice; it has no capabilities for file writing, system command execution, or network communication.
Sanitization: No sanitization or validation of the ingested context files is performed.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill reads local project files to customize its recommendations. This data access is targeted at project-specific marketing information and does not attempt to access sensitive system files (e.g., SSH keys, AWS credentials) or exfiltrate data to external servers.

pricing