senior-software-architect
Fail
Audited by Gen Agent Trust Hub on Mar 31, 2026
Risk Level: CRITICALDATA_EXFILTRATIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: Automated security scanners have identified a malicious URL within the technical documentation. The domain "https://api.talent-architect.com", which is used in the skill's TypeScript configuration examples, is flagged as a phishing risk.
- [CREDENTIALS_UNSAFE]: The skill explicitly documents existing architectural debt where sensitive API keys have been stored in the ".env" files within the repository (Section 13). Although it mentions rotating them, the reference to secrets in code repositories indicates a historical and ongoing risk of credential exposure.
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its core functionality of processing untrusted external data (CVs and job descriptions) using AI models. The instructions provide prompt templates that interpolate this untrusted content directly into messages (Section 14) without boundary markers or instructions to ignore embedded commands.
- Ingestion points: Functions such as
parseCv(file)andanalyzeAts(cv, jobDesc)ingest untrusted data from external uploads. - Boundary markers: The provided prompt templates lack delimiters (e.g., XML tags or clear separators) and do not include instructions for the AI to disregard instructions contained within the user-provided data.
- Capability inventory: The skill has access to network operations (OpenRouter/Anthropic APIs), database access (Firestore), and file system parsing capabilities.
- Sanitization: While the skill emphasizes Zod for schema validation, it lacks any sanitization mechanisms specifically designed to prevent adversarial text from influencing AI behavior.
Recommendations
- AI detected serious security threats
- Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata