pentest-validation
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes responses from external, potentially untrusted target URLs to validate vulnerability findings. \n
- Ingestion points: The agent context receives data from server responses via the
target_urlduring the validation pipeline (SKILL.md). \n - Boundary markers: Although the skill uses structured task orchestration, it lacks explicit instructions or delimiters designed to isolate and ignore embedded LLM instructions within the target site's response data (SKILL.md). \n
- Capability inventory: The skill has broad capabilities including performing network requests, reading project source code, and generating exploitation payloads (SKILL.md). \n
- Sanitization: The provided documentation does not describe any specific sanitization or filtering processes to handle external content before it is processed by the agent's reasoning engine.
Audit Metadata