agent-bom ingest
Validate and import cloud resource inventory JSON to generate local audit reports
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 ingest. It is used for: Validate and import cloud resource inventory JSON to generate local audit reports Full Skill content: https://321skill.com/skills/agent-bom-ingest-x-4/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 pain point of 'scanning without credentials' in cloud resource inventory and security audits. In real-world operations, operators often have already exported inventory JSON from platforms like AWS, Azure, GCP, or Snowflake using their own scripts or third-party tools, but lack a standardized tool for validation and local analysis. agent-bom-ingest is built precisely for this purpose. It directly accepts these JSON files, performs format validation, and executes local scanning, generates dependency graphs, conducts policy audits, and exports reports—all without exposing cloud credentials externally.
Usage is straightforward: you only need to prepare an inventory JSON file compliant with the agent-bom specification (which can be obtained from an operator's own collector, CMDB export, or AI Agent workflow). You can then trigger validation, scanning, and export processes using natural language commands, such as "Validate inventory.json and scan for vulnerabilities" or "Export the results in SARIF format for CI pipeline use." All operations are based on the local MCP protocol, keeping data within the domain, with an optional push to the operator's own control plane.
It is ideal for teams or individuals who need to regularly inventory cloud resource assets and perform security and compliance audits. This is especially useful for operations and security engineers who already have mature inventory collection processes but want to obtain localized, traceable audit reports and dependency graphs without adding extra credentials. For cloud migration projects, multi-cloud environment governance, and regulatory compliance inspection scenarios, this Skill can significantly reduce toolchain complexity.
It is recommended to use this tool uniformly after establishing a standardized inventory JSON output process. Note that it primarily relies on the local agent-bom engine (0.84.4+) and requires the input JSON to strictly adhere to the built-in inventory.schema.json contract. If the inventory source does not meet the schema, format conversion via an adapter is necessary first. Additionally, all network requests (e.g., vulnerability database queries) are optional and are completed locally by default.
Key Features
Unlike native cloud scanning tools such as AWS Config or Azure Policy, agent-bom-ingest requires no cloud credentials or API calls. It operates entirely offline based on the inventory JSON provided by the operator, eliminating security risks associated with credential exposure while supporting unified audits across hybrid multi-cloud environments.
Limitations
Requires Python 3.11+ and agent-bom ≥0.84.4. The input inventory JSON must strictly adhere to the built-in inventory.schema.json format; otherwise, pre-conversion is necessary.
FAQ
Are cloud platform access keys required?
No. The tool only processes JSON files provided by the operator and does not require any cloud credentials. Optional control plane push requires an API Token from the operator but does not actively fetch credentials.
Which cloud platforms are supported?
Supports AWS, Azure, GCP, Snowflake, as well as JSON generated by CMDB and endpoint collectors. Any format compliant with the schema is acceptable.
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-ingest-x-4/raw/index.md to read the original Skill definition (Markdown format) for agent-bom ingest, and install it according to the instructions.
Raw Markdown URL for AI: /skills/agent-bom-ingest-x-4/raw/index.md