oma-architecture
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the local codebase through search and read operations, which could contain malicious instructions designed to influence the agent's architectural recommendations.
- Ingestion points: The agent is instructed to use
rg(ripgrep) to search the codebase for architectural keywords and read relevant source files as documented inSKILL.mdandresources/execution-protocol.md. - Boundary markers: The instructions do not specify the use of delimiters or warnings to the model to ignore embedded instructions within the source code being analyzed.
- Capability inventory: The skill has the capability to write architectural artifacts (ADRs, recommendations) to the local filesystem under the
.agents/results/architecture/directory. - Sanitization: There are no explicit instructions for the agent to sanitize, escape, or filter content retrieved from the external files before including it in reasoning or final outputs.
Audit Metadata