outreach-specialist-skill-founder
Pass
Audited by Gen Agent Trust Hub on Jul 30, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The instructions utilize authoritative directives (e.g., 'BLOCKING REQUIREMENT', 'MANDATORY') to ensure the agent correctly follows the execution logic and reads necessary reference files. These instructions are aimed at task accuracy and do not attempt to bypass core safety guardrails.
- [DATA_EXFILTRATION]: The skill is designed to read business context from a local file named
FOUNDER_CONTEXT.mdto personalize outreach. This data is used exclusively for content generation within the user's session and there is no evidence of attempts to exfiltrate this information to external endpoints. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from user inputs and the
FOUNDER_CONTEXT.mdfile, creating a potential attack surface. - Ingestion points:
$ARGUMENTS, user messages, and theFOUNDER_CONTEXT.mdfile (SKILL.md). - Boundary markers: The skill does not implement specific delimiters or 'ignore' instructions for external data.
- Capability inventory: The agent's capabilities are limited to reading local files (
Read) and asking the user questions (AskUserQuestion). - Sanitization: No explicit sanitization or filtering of external data is present.
- [SAFE]: The skill does not contain any obfuscated code, hidden URLs, remote code execution patterns, or persistence mechanisms. All referenced files are local to the skill package.
Audit Metadata