architectural-analysis
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONMETADATA_POISONING
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted codebase content which could contain malicious instructions designed to manipulate the agent's analysis or the content of the generated report.
- Ingestion points: All source code files (.ts, .tsx, .js, .jsx) processed during steps 2 through 6 of the audit in
SKILL.md. - Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded prompts within the source code being analyzed, increasing the risk that the agent may follow instructions found inside comments or strings.
- Capability inventory: The skill uses standard shell tools (
find,grep,wc,sort,head) for analysis and has the capability to write a report file to the local.audits/directory. - Sanitization: There is no explicit sanitization of codebase content before it is interpolated into the markdown report or summary.
- [METADATA_POISONING]: The skill's metadata identifies the author as 'Pedro Nauck' and provides links to his personal GitHub repository, which contradicts the provided author context of 'marcioaltoe'. This discrepancy is deceptive and could mislead users or administrators about the skill's origin and maintenance.
Audit Metadata