redis
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONPRIVILEGE_ESCALATIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides numerous shell commands for package management (apt, dnf), service control (systemctl), and database interaction (redis-cli) to configure and operate the Redis server.
- [PRIVILEGE_ESCALATION]: The skill requires sudo privileges for the installation of the Redis server and for modifying system-level configuration files such as /etc/redis/redis.conf and /etc/redis/sentinel.conf.
- [EXTERNAL_DOWNLOADS]: The Docker Compose configuration fetches the official redis:7-alpine image and the well-known rediscommander/redis-commander image from Docker Hub.
- [INDIRECT_PROMPT_INJECTION]: The skill enables an AI agent to interact with an external data store (Redis), which creates a surface for indirect prompt injection if the agent reads and processes untrusted data. * Ingestion points: Redis read operations including GET, HGETALL, RPOP, SUBSCRIBE, and monitoring commands like INFO, SLOWLOG, and MONITOR (found in SKILL.md). * Boundary markers: None provided in the instructions to help the agent distinguish between data and instructions. * Capability inventory: System-level configuration access via sudo, file modification of service configurations, and full database administrative capabilities via redis-cli (found in SKILL.md). * Sanitization: The skill does not provide instructions for escaping or validating data retrieved from the Redis store before processing.
Audit Metadata