migrate-to-shoehorn
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute
npm ifor package installation andgrepto recursively search for type assertion patterns (as [A-Z]) within local test files. - [EXTERNAL_DOWNLOADS]: Fetches the
@total-typescript/shoehornutility from the npm registry, which is the intended purpose of the migration skill. - [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect injection as it ingests and processes local test file content to perform migrations.
- Ingestion points: Local test files identified by the
grepcommand (specified inSKILL.md). - Boundary markers: None explicitly defined to separate the code being migrated from the agent's instructions.
- Capability inventory: The agent has capabilities to execute shell commands (
npm,grep) and modify local files as part of the migration workflow (specified inSKILL.md). - Sanitization: No specific sanitization or validation of the source code content is described before processing.
Audit Metadata