routeros-quickchr-cli
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill requires executing local command-line tools to manage virtual machine lifecycles and execute RouterOS commands. Evidence includes usage of
quickchr(add, start, exec, stop, remove) andcentrs(retrieve, execute) as described throughout SKILL.md and the grounding logs. - [INDIRECT_PROMPT_INJECTION]: The skill processes output from RouterOS instances, creating a surface where malicious data could influence agent behavior.
- Ingestion points: Data enters the context via the output of
centrs retrieveandquickchr execcalls in SKILL.md and references/cli-grounding.md. - Boundary markers: No delimiters or explicit instructions are provided to the agent to treat machine output as untrusted data.
- Capability inventory: The skill enables shell command execution on the host and network interaction with the guest VM APIs.
- Sanitization: There is no evidence of sanitization or filtering of the content returned by the RouterOS commands.
- [EXTERNAL_DOWNLOADS]: The skill documentation includes commands for installing the required CLI tools from the npm registry.
- Evidence:
bun add @tikoci/quickchr@0.4.8andbun add @tikoci/centrsin the references/cli-grounding.md file. These packages are authored by the skill vendor.
Audit Metadata