flux-opencode-scope
Warn
Audited by Gen Agent Trust Hub on Jul 31, 2026
Risk Level: MEDIUMCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill dynamically assembles shell commands for the
fluxctlutility by interpolating variables like<Short title from problem statement>and<Task title>into command strings. Because these values are sourced from untrusted user input and specification documents, there is a risk of command injection if the agent does not robustly escape shell metacharacters (e.g.,;,&,|) before execution. - [EXTERNAL_DOWNLOADS]: The workflow incorporates 'scouts' (such as
flux:docs-scoutandflux:github-scout) that retrieve data from external sources like GitHub and online documentation. This ingestion of external content is a documented feature for researching project context, but it establishes a surface for indirect prompt injection where malicious instructions embedded in the source data could influence the agent's behavior.
Audit Metadata