skillsman-init
Pass
Audited by Gen Agent Trust Hub on May 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill is designed to execute the
skillsmancommand-line tool to initialize and manage agent skills. It includes a specific instruction to fallback to a hardcoded local path (/Users/yatmn/Projects/skillsman/bin/skillsman) if the binary is not in the system PATH, which matches the author's namespace. - [EXTERNAL_DOWNLOADS]: The
skillsman initandskillsman addcommands facilitate the downloading and installation of skill configurations from external sources into the project directory as a core part of its functionality. - [PROMPT_INJECTION]: The skill's workflow involves reading local repository files to automate project type detection (Indirect Prompt Injection surface). This creates a potential vector where a malicious repository could influence the agent's choice of installation profiles or command arguments. This is mitigated by the skill's requirement for explicit user confirmation before execution.
- Ingestion points: Step 2 in
SKILL.md(inspecting repository files). - Boundary markers: Absent for the file inspection phase.
- Capability inventory: Execution of shell commands (
skillsman show/init/add/doctor). - Sanitization: No explicit sanitization or filtering mentioned for repository content used for project typing.
Audit Metadata