pulumi-cli

Pass

Audited by Gen Agent Trust Hub on Mar 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill documents various Pulumi CLI commands (e.g., pulumi up, pulumi stack, pulumi state) that perform administrative operations on cloud infrastructure and local file systems.
  • [EXTERNAL_DOWNLOADS]: Mentions the ability to download and initialize projects from external templates and Git repositories (e.g., github.com/pulumi/templates).
  • [REMOTE_CODE_EXECUTION]: Pulumi's core functionality involves executing infrastructure-as-code programs and policy packs, which can be retrieved from local or remote sources.
  • [CREDENTIALS_UNSAFE]: Documents the use of sensitive environment variables for authentication, such as PULUMI_ACCESS_TOKEN, AWS_ACCESS_KEY_ID, and ARM_CLIENT_SECRET. These are presented as setup instructions using standard placeholders.
  • [PROMPT_INJECTION]: The skill identifies surfaces for indirect prompt injection, as it documents commands that ingest external data from files like backup.json, resources.json, and config.json while maintaining significant capabilities to manage cloud resources.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 26, 2026, 03:44 PM