personal statement
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from user-provided resumes, transcripts, and application prompts, which could theoretically contain instructions to override the agent's behavior.\n
- Ingestion points: Stage 1 (Gather) and Stage 4 (Polish) prompt users to provide CVs, transcripts, and text dumps, and the agent reads existing context from
.claude/personal-statement-context.md.\n - Boundary markers: The instructions do not define specific delimiters or security warnings to isolate user-supplied text from the agent's core logic.\n
- Capability inventory: The skill is configured to use tools like
create_fileandreplace_string_in_fileto manage local documentation.\n - Sanitization: No input validation or content filtering is performed on the provided documents before they are used to generate the personal statement.\n- [SAFE]: The skill's use of local persistence via the
.claude/directory is a functional feature for maintaining user state during the multi-step writing process and does not involve unauthorized system access or external network calls.
Audit Metadata