The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for executing shell commands and MongoDB shell (mongosh) scripts to manage databases, create users, configure replica sets, and perform backups.
[PRIVILEGE_ESCALATION]: Includes administrative tasks requiring sudo for package installation via apt, managing system services with systemctl, and writing to protected directories such as /etc/apt/ and /etc/mongodb-keyfile.
[EXTERNAL_DOWNLOADS]: Fetches official GPG keys and repository configurations from MongoDB's official domains (mongodb.org and repo.mongodb.org) during the installation process.
[INDIRECT_PROMPT_INJECTION]: The skill documents methods to query and aggregate data from MongoDB collections. If an agent uses these capabilities to process untrusted data stored in a database, it creates a potential surface for injection attacks.
Ingestion points: Database queries such as db.users.find() and complex aggregation pipelines.
Boundary markers: No explicit delimitation or warnings to ignore embedded instructions in retrieved data are provided.
Capability inventory: Significant capabilities including root-level command execution (sudo), service management, and file system operations (mongodump).
Sanitization: The documentation does not include specific sanitization or validation steps for data retrieved from the database before it is processed by the agent.

mongodb