floci
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill provides scripts and instructions to install the floci CLI by downloading and running an installer from the official developer domain (floci.io). This is standard installation behavior for this toolset.
- [DYNAMIC_EXECUTION]: Instructions in SKILL.md and the scripts/ensure-running.sh utility use the eval command to process the output of floci env. This is a common and necessary pattern used to export required environment variables into the user's current shell session.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to fetch and process external content, which represents a potential indirect prompt injection surface.
- Ingestion points: Documentation and configuration files are fetched from floci.io and the floci-io GitHub organization.
- Boundary markers: External content is ingested directly without specific isolation markers.
- Capability inventory: The skill can execute local commands via the floci and docker CLIs and use eval for shell configuration.
- Sanitization: Input is processed as raw Markdown or parsed from tool-generated JSON/text.
- [COMMAND_EXECUTION]: The skill executes various local commands using the floci and docker CLIs to manage the lifecycle of cloud emulators. This is the primary intended function of the skill.
Audit Metadata