The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/agent.py utilizes the subprocess.run method to execute the cartography command-line utility. This is a standard wrapper pattern to invoke the tool for asset synchronization using parameters provided via the script's own CLI.
[PERSISTENCE]: The SKILL.md file provides instructions for configuring a cron job to automate regular cloud asset synchronization. This use of a persistence mechanism is a legitimate and standard method for maintaining up-to-date security inventory data.
[EXTERNAL_DOWNLOADS]: The skill directs users to install the cartography and neo4j packages from the official Python Package Index (PyPI) and uses standard Neo4j Docker images. These resources are from well-known and reputable sources consistent with the skill's purpose.
[INDIRECT_PROMPT_INJECTION]: The skill processes structured metadata from cloud environments via the Cartography sync process and subsequent graph queries.
Ingestion points: Infrastructure metadata is fetched from Cloud APIs (AWS, GCP, Azure) and stored in a Neo4j database.
Boundary markers: Interaction with the data is performed through hardcoded and structured Cypher queries within the scripts/agent.py and scripts/process.py files.
Capability inventory: The skill has the ability to execute the cartography tool and write report files to the local disk.
Sanitization: The results are handled as structured data records by the Neo4j driver and the Python scripts, minimizing risks associated with untrusted data processing.

performing-cloud-asset-inventory-with-cartography