conductor-new-track
Pass
Audited by Gen Agent Trust Hub on Jul 22, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of supplementary capabilities by downloading
SKILL.mdfiles from a defined catalog inassets/catalog.md. These downloads target the author's own repositories and official Firebase repositories. All downloads require explicit user confirmation via a multiple-choice prompt before execution. - [COMMAND_EXECUTION]: Shell commands are utilized for project organization, including creating directories (
mkdir -p), fetching documentation (curl -sSL), and managing version control (git commit). These operations are restricted to the local project environment (.agents/skills/,conductor/). - [PROMPT_INJECTION]: The skill implements a 'Sequential Questioning' protocol and strict state verification. It mandates asking questions one at a time and waiting for user input, which reduces the risk of the agent bypassing instruction sets or making unauthorized assumptions during complex planning phases.
Audit Metadata