The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to query and aggregate data from MongoDB collections, which serves as a potential vector for indirect prompt injection if the database content is attacker-controlled.
Ingestion points: Data enters the agent's context through db.users.find(), db.orders.aggregate(), and db.system.profile.find() as documented in SKILL.md.
Boundary markers: The instructions do not specify any delimiters or safety prompts to prevent the agent from following instructions embedded within the retrieved data.
Capability inventory: The skill provides access to powerful system and database tools including sudo, apt, systemctl, mongosh, mongodump, and mongorestore (SKILL.md).
Sanitization: There are no instructions for sanitizing, filtering, or validating the database content before it is processed by the agent.
[PRIVILEGE_ESCALATION]: The skill requires and uses administrative privileges (sudo) for critical tasks such as package installation (apt install), system service management (systemctl), and modifying system configuration files (/etc/apt/sources.list.d/).
[EXTERNAL_DOWNLOADS]: Fetches the official GPG security key and repository configuration from MongoDB's official domains (mongodb.org) to facilitate the installation of the database software.
[COMMAND_EXECUTION]: Employs various shell commands and MongoDB shell (mongosh) scripts to perform database administration, user management, and performance monitoring.

mongodb