reset-cursor-acp
Warn
Audited by Gen Agent Trust Hub on Oct 20, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPRIVILEGE_ESCALATION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a bash script
scripts/reset-cursor-acp.shwhich contains several system-level commands. - Evidence: Use of
ps,grep,kill -TERM, andkill -KILLto managecursor-agentprocesses. - Evidence: Execution of the
bbCLI andcursor-agentbinary to check status and versions. - [PRIVILEGE_ESCALATION]: The skill mentions that if sandboxing blocks
ps, the user should rerun the script outside the sandbox, which is a recommendation to lower security boundaries. - Evidence:
SKILL.mdtext: "If sandboxing blocks ps with 'operation not permitted', rerun the script outside the sandbox." - [REMOTE_CODE_EXECUTION]: The script uses
python3 -cfor inline processing of JSON data from a command-line tool, which is a form of dynamic execution but targets local data. - Evidence:
SKILL.mdcontains a snippet usingpython3 -c 'import json,sys; ...'to parse thread IDs. - [SAFE]: The network operations mentioned (
cursor-agent login,bb updates) are part of the intended functionality for the Cursor and bb tools and do not involve untrusted third-party domains or sensitive data exfiltration.
Audit Metadata