api-architect
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill architecture creates a surface for indirect prompt injection by design. It requires the agent to process and analyze external, potentially untrusted documents including business requirements, workflow evidence, API specifications, schemas, and source code (Section 12.1, 12.3). The instructions lack explicit guidance for the agent to use boundary markers or to ignore embedded instructions within these ingested artifacts, which could allow a malicious file to subvert the agent's governance role.
- Ingestion points: File SKILL.md, Section 12.1 (Requirements, workflow evidence, domain language) and Section 12.3 (Markdown, structured records, API specifications, schemas, code).
- Boundary markers: Absent. The skill does not instruct the agent to use specific delimiters or protective prompting techniques when processing external content.
- Capability inventory: The agent has the capability to generate source code, API contracts, and Architecture Decision Records (ADRs) (Sections 13 and 18).
- Sanitization: Absent. No filtering or validation logic is specified for the content of ingested artifacts.
Audit Metadata