mitre-attack-lookup
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill clones a remote repository from GitHub (https://github.com/woohyun212/mitrize.git) which belongs to the skill author. This repository contains the ATT&CK dataset in Markdown format and utility scripts.
- [REMOTE_CODE_EXECUTION]: The workflow involves executing a Python script (scripts/query_attack_md.py) that is downloaded as part of the author's mitrize repository to perform queries.
- [COMMAND_EXECUTION]: The skill utilizes several standard shell commands including git (for cloning and updating), python3 (for running the query script), and filesystem tools like cat, grep, and ls to interact with the downloaded data.
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from the external mitrize repository which could theoretically contain malicious instructions. 1. Ingestion points: Markdown files within the mitrize repository (SKILL.md). 2. Boundary markers: Absent. 3. Capability inventory: Shell command execution via git, python3, cat, and grep (SKILL.md). 4. Sanitization: Absent.
Audit Metadata