cheat-engine-pointer-chains
Warn
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install an external package,
game-re-sigtool, from the PyPI registry. This package is a third-party, non-standard dependency.- [COMMAND_EXECUTION]: The workflow relies on the execution of shell commands through thesigtoolcommand-line utility to perform memory analysis and verify pointer chains.- [DYNAMIC_EXECUTION]: Thesigtool practicecommand generates and writes binary executable files (described as a "synthetic game") to the local file system.- [PRIVILEGE_ESCALATION]: The documentation indicates that thesigtoolutility may require administrative (elevated) privileges to successfully access and read the memory of other processes (PROCESS_VM_READ).- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data directly from external process memory at runtime. This provides a potential attack surface for indirect prompt injection if the memory being read contains attacker-controlled strings or instructions. - Ingestion points: External process memory accessed via the
sigtool chaincommand inSKILL.md. - Boundary markers: There are no explicit delimiters or boundary markers defined for the data returned from memory.
- Capability inventory: The skill possesses shell command execution capabilities (
sigtool) and the ability to install packages (pip). - Sanitization: No sanitization, validation, or filtering of the data retrieved from process memory is described in the workflow.
Audit Metadata