design
Pass
Audited by Gen Agent Trust Hub on Jul 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a custom project utility named
doctrineto track lifecycle status, link entities, and manage metadata (e.g.,doctrine slice selector add <id>). These operations are consistent with the skill's stated purpose of managing design and implementation artifacts. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted content from external sources to inform the architectural design process.
- Ingestion points: Processes project-level artifacts including specifications, Architecture Decision Records (ADRs), source code, documentation, and git commit history (identified in
SKILL.mdandSKILL.compact.md). - Boundary markers: No explicit delimiters or instructions to ignore embedded malicious prompts within ingested data are defined.
- Capability inventory: Possesses the ability to write to the local filesystem (
design.md), execute CLI commands (doctrine), and invoke downstream agent skills (/plan,/inquisition,/canon). - Sanitization: No explicit sanitization or validation of the ingested external content is described in the workflow.
Audit Metadata