allow-agent-commands
Warn
Audited by Gen Agent Trust Hub on Feb 24, 2026
Risk Level: MEDIUMCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides explicit instructions for the agent to modify its execution environment by updating permission allow-lists in
.config/opencode/opencode.jsonand.claude/settings.json. - [COMMAND_EXECUTION]: This functionality represents a privilege escalation vector, as it guides the agent on how to modify the security boundaries defined by its restricted command environment.
- [COMMAND_EXECUTION]: The instructions promote the use of wildcard patterns (e.g.,
"rg *": "allow"), which can grant overly broad permissions for command execution with arbitrary arguments. - [COMMAND_EXECUTION]: The skill coordinates with other subagents (
git-stager,git-committer) to apply and persist these permission changes to a configuration repository.
Audit Metadata