create-group
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied input such as group titles, descriptions, and usernames which are passed directly to a CLI tool. If the underlying script does not properly sanitize these inputs, it could lead to command injection or logic subversion.
- Ingestion points: User-provided strings for
title,--about, and--usersparameters used in thebuncommand. - Boundary markers: Absent. No delimiters or instructions to ignore embedded commands in the user input are provided.
- Capability inventory: Executes shell commands via
bunand accesses local filesystem paths. - Sanitization: None visible in the instruction file.
- [DATA_EXFILTRATION]: The skill explicitly references and verifies the existence of Telegram session files located at
~/.local/share/gramjs/<profile>.session. While this is necessary for Telegram API interactions, these files contain highly sensitive credentials (auth keys) that could be targeted for exfiltration if the environment is compromised.
Audit Metadata