reset-cursor-acp

Pass

Audited by Gen Agent Trust Hub on Sep 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses a bash script and a Python one-liner to manage local system processes and interact with the bb CLI. It uses ps and kill to identify and terminate specific cursor-agent subprocesses and executes bb thread stop to release runtimes.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes and displays thread metadata including user-provided titles, creating a potential surface for instructions embedded in data to influence agent behavior.
  • Ingestion points: The Python snippet in SKILL.md and the ps parsing logic in scripts/reset-cursor-acp.sh read untrusted thread information.
  • Boundary markers: No explicit delimiters or 'ignore embedded instructions' warnings are present in the output displayed to the agent.
  • Capability inventory: The scripts/reset-cursor-acp.sh script has capabilities to terminate processes (kill) and stop threads (bb thread stop).
  • Sanitization: No sanitization or validation of the thread data is performed before it is processed or displayed.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 5, 2026, 08:17 PM
Security Audit — agent-trust-hub — reset-cursor-acp