stelow-product-interface-alternatives
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill uses standard shell utilities such as
ls,grep, andawkto manage local workflow files. For example, it usesgrepto extract appetite settings from project specifications andawkto isolate specific proposal sections for user selection. - [REMOTE_CODE_EXECUTION]: The skill dynamically sources a configuration script from a relative path (
../../stelow-product-orchestrator/references/cli-tools/read-config.sh). This script is part of the vendor's own project ecosystem and is used to read global settings. - [DATA_EXPOSURE]: The skill reads local project files including
spec-product.mdandtech-recon.mdto provide context to subagents. This behavior is confined to the local project directory and follows the intended product design workflow. - [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: Reads external data from
spec-product.mdandtech-recon.md(inSKILL.md). - Boundary markers: None explicitly used during the initial file read, although subagents are given specific task contexts.
- Capability inventory: Uses
subagentsandvisual_reviewtools, and performs file-write operations viaawkand bash redirection (inSKILL.md). - Sanitization: No specific sanitization or escaping of the ingested file content is performed before passing it to subagents.
Audit Metadata