The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/agent.py script executes system commands via subprocess.run to interact with gcloud and gsutil. These calls are used for legitimate discovery tasks such as listing buckets, retrieving IAM policies, and auditing service account configurations.
[EXTERNAL_DOWNLOADS]: The documentation in references/api-reference.md provides instructions to download the GCPBucketBrute security tool from Rhino Security Labs' official GitHub repository.
[INDIRECT_PROMPT_INJECTION]: The skill processes output from external cloud resources (bucket listings and IAM policies) which could potentially contain adversarial content intended to influence downstream agent behavior.
Ingestion points: Data enters the environment via subprocess.run calls to gsutil ls and gcloud projects get-iam-policy in scripts/agent.py.
Boundary markers: None are present in the script to distinguish between system output and potentially malicious cloud resource names.
Capability inventory: The skill has the capability to execute shell commands via subprocess.run and write to the local file system using json.dump.
Sanitization: While command output is truncated for display, no validation or escaping is performed on the data retrieved from the GCP environment before it is stored or processed.

performing-gcp-penetration-testing-with-gcpbucketbrute