secops-scanner

Warn

Audited by Gen Agent Trust Hub on Oct 17, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes multiple docker run commands to initiate scans using Semgrep, OSV-Scanner, Trivy, and Gitleaks. These commands are intended to pull the latest images and scan the project source code mounted as a read-only volume.
  • [PRIVILEGE_ESCALATION]: For container image scans, the skill mounts the host's Docker socket (/var/run/docker.sock) into the Trivy container as described in references/container-and-iac-trivy.md. This configuration provides the container with full control over the host's Docker daemon, which can be exploited for container escape and host-level root access if the containerized tool or the images it processes are compromised.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes raw outputs from external security scanners, making it vulnerable to malicious content embedded in findings that could influence the agent's behavior. 1. Ingestion points: Scan results from Semgrep, OSV-Scanner, Trivy, and Gitleaks are processed to generate SECURITY-AUDIT-REPORT.md. 2. Boundary markers: No explicit delimiters or warnings to ignore embedded instructions are present during output processing. 3. Capability inventory: run_command, write_to_file, and replace_file_content are available to the agent (SKILL.md). 4. Sanitization: No sanitization or validation of external scanner content is observed before report generation.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Oct 17, 2026, 11:08 AM