The Agent Skills Directory

[SAFE]: The skill consists of Markdown documentation and templates for generating project proposals. It does not contain executable code, binary files, or installation scripts.
[DATA_EXFILTRATION]: No hardcoded credentials, sensitive file path access, or unauthorized network operations were detected. File system operations are limited to organizing audit results within a local directory.
[PROMPT_INJECTION]: The skill processes untrusted content from external website audits, which is an attack surface for indirect prompt injection. 1. Ingestion points: Prospect website URLs and content extracted during audits. 2. Boundary markers: Absent in the provided templates. 3. Capability inventory: File system writes to the 'proposal-data/' directory. 4. Sanitization: No specific data validation or escaping is defined in the instructions. This surface is inherent to the skill's intended functionality and does not involve critical capabilities like arbitrary command execution.

client-proposal