speckit-preset
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill documents the use of the
specifycommand-line tool for preset management, including lifecycle commands likeadd,remove,enable, anddisable.- [EXTERNAL_DOWNLOADS]: The skill provides instructions for downloading content from external sources through commands such asspecify preset add --from <url>andspecify preset catalog add <url>.- [INDIRECT_PROMPT_INJECTION]: The skill identifies an attack surface for indirect prompt injection by facilitating the ingestion of remote presets and catalogs. - Ingestion points: External content is pulled into the agent's context via the
--from <url>andcatalog add <url>arguments inSKILL.md. - Boundary markers: The skill does not define specific prompt boundary markers, but it does include a security warning advising users to only permit installation (
--install-allowed) for trusted catalogs. - Capability inventory: The skill leverages shell execution capabilities through the
specifyCLI. - Sanitization: No explicit sanitization or validation of the remote content is described within the instruction file, relying on the underlying CLI implementation.
Audit Metadata