agent-bom discover gcp
Discover AI Agent and MCP Asset Inventory on GCP
Install & Use
Copy this prompt and send it to your AI assistant (Claude / Cursor / TRAE / Codex / WorkBuddy etc.) to auto-install:
Help me install this AI Skill: agent-bom discover gcp. It is used for: Discover AI Agent and MCP Asset Inventory on GCP Full Skill content: https://321skill.com/skills/agent-bom-discover-gcp-x-7/raw/index.md Read that page and install it.
The prompt includes a link to the full Skill content. You can also view the full content.
This Skill addresses the challenge of systematically inventorying AI Agent and MCP-related assets in GCP environments. In real-world operations, developers and SREs need a comprehensive view of resource usage across services like Vertex AI, Cloud Run, Cloud Functions, and GKE. However, manually querying disparate service APIs is time-consuming and prone to missing critical assets. With a single command, this tool automatically discovers and generates a standardized agent-bom inventory JSON, providing a reliable data foundation for subsequent asset scanning, compliance audits, and cost analysis.
Usage is straightforward: simply run the provided Python script in an environment with GCP read-only permissions, specifying the project ID and region. The script leverages the GCP SDK (using Application Default Credentials, Workload Identity, or short-lived service accounts) to make read-only API calls, automatically cataloging all AI Agent and MCP-related resources, and finally outputting a normalized inventory JSON file. The entire process does not require passing long-term credentials to agent-bom; credentials remain in the operator's local environment, ensuring security.
It is ideal for teams or individuals who need to regularly inventory GCP AI infrastructure, especially SREs, cloud architects, and agent CI/CD developers. For users already employing agent-bom for asset management, this Skill integrates seamlessly into existing workflows, directly generating inventory files ready for subsequent scanning. Furthermore, in large organizations with multiple projects and regions, this automation capability significantly reduces manual inventory costs and improves asset visibility.
It is recommended to integrate this tool into unified CI/CD or asset management processes as a regular part of daily operations. Note that the tool performs read-only discovery operations only and will not modify any GCP resources. It requires Python 3.11+ and that the operator has GCP read-only permissions. The output file path is manually specified by the operator to avoid accidentally writing sensitive data to uncontrolled locations. Before first use, ensure local credentials are properly configured (e.g., via gcloud auth application-default login).
Key Features
Unlike manually querying with gcloud commands, this tool is specifically designed for AI Agent and MCP assets, automatically outputting a standardized agent-bom format that can be directly used for subsequent vulnerability scanning, compliance checks, and cost analysis. Additionally, it does not require passing long-term GCP credentials to agent-bom, mitigating credential leakage risks—a common requirement in other similar tools that often necessitate hardcoded service account keys.
Limitations
Requires Python 3.11+, depends on the GCP SDK (ADC, Workload Identity, or scoped service account), supports only the GCP platform, and performs read-only discovery operations only—it does not support modifying or configuring resources.
FAQ
What GCP permissions are required to run this?
Only GCP read-only permissions are needed, achievable via Application Default Credentials, Workload Identity, or short-lived service accounts. No long-term credentials are required.
Where is the output file saved?
The output path is specified by the operator when executing the command. The script writes the normalized agent-bom inventory JSON to that path and does not automatically upload it to any external service.
Installation guide for AI assistants
If your AI coding assistant (Claude Code, Cursor, TRAE etc.) can see this page, send it this message to auto-install:
Visit https://321skill.com/skills/agent-bom-discover-gcp-x-7/raw/index.md to read the original Skill definition (Markdown format) for agent-bom discover gcp, and install it according to the instructions.
Raw Markdown URL for AI: /skills/agent-bom-discover-gcp-x-7/raw/index.md