cloud-logging-cross-project-configuration
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- Command Execution (IAM and Resource Management): The skill executes commands to modify cloud infrastructure and security policies. While these are necessary for the skill's primary purpose, they involve high-risk operations such as granting IAM roles and deleting log buckets. * Evidence: Use of
gcloud projects add-iam-policy-bindingandgcloud logging buckets deleteinSKILL.md. * Context: The skill mitigates this risk by implementing mandatory confirmation tiers (Tier B and Tier D) that require explicit user approval before sensitive commands are run. - Indirect Prompt Injection Surface: The skill ingests log data from external sources, which could potentially contain instructions designed to influence the agent's behavior. * Ingestion points:
gcloud logging readinSKILL.mdallows the agent to pull log content into its context. * Boundary markers: There are no explicit delimiters or warnings to ignore instructions within the ingested log data. * Capability inventory: The skill has the ability to execute subprocesses for IAM modification (gcloud projects add-iam-policy-binding) and resource deletion (gcloud logging buckets delete). * Sanitization: No sanitization or filtering of the retrieved log content is specified.
Audit Metadata