setup-brad-frost-skills
Pass
Audited by Gen Agent Trust Hub on Oct 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to scan specific local directories to verify the installation status of other tools, including paths such as
~/.claude/skills/,.claude/skills/,~/.agents/skills/, and.agents/skills/. It also checks for the existence of a configuration file at~/.config/ai-limits/configand recommends a manual installation command (npx skills add bradfrost/skills -g) which targets the author's own package. - [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests data from external, potentially untrusted sources during its configuration phase.
- Ingestion points: The instructions specify reading component vocabulary and runtime information from external manifests, MCP (Model Context Protocol) servers, or code repositories (documented in
SKILL.mdsection 3). - Boundary markers: The instructions do not provide clear delimiters or "ignore embedded instructions" warnings for the agent when processing this external content.
- Capability inventory: The skill has the capability to transition into and execute other instructional scripts (e.g.,
limits-setup,intake/PROFILE-INTAKE.md,intake/SYSTEM-INTAKE.md,intake/STORYBOOK-INTAKE.md) based on the data retrieved from these external sources. - Sanitization: No specific sanitization, escaping, or validation of the external content is described before it is interpolated into the agent's context.
Audit Metadata