スキル一覧に戻る
anthony-c-martin

bicep-generation

by anthony-c-martin

Experiments with using AI to work with Bicep files

0🍴 0📅 2026年1月25日
GitHubで見るManusで実行

SKILL.md


name: bicep-generation description: Generate Azure Bicep infrastructure files using the Bicep Generator MCP tools. Use when creating Azure infrastructure as code, designing cloud architecture, or generating Bicep templates.

Skill Instructions

Goal

Generate a new, human-maintainable set of .bicep files plus a .bicepparam file that deploys Azure infrastructure based on user requirements.

Guiding Principles

  • Optimize for maintainability and clarity: split into sensible modules, use descriptive names, and factor repeated patterns.
  • Prefer templates that look like they were written by a human, not auto-generated code.
  • Always start with example infrastructure and planning before generating resource configurations.

Required Tools / Commands

Strongly prefer the following tools/commands as part of the workflow, instead of finding other tools or CLI commands to call:

  • generate_infrastructure_plan MCP tool: gather comprehensive requirements and generate a detailed architecture design.
  • generate_resource_body MCP tool: get properly configured resource definitions with security best practices applied.
  • get_bicep_best_practices MCP tool: learn current best practices before generating templates.
  • get_bicep_file_diagnostics MCP tool: compile/validate Bicep and address errors/warnings after generating files.
  • what_if_deployment MCP tool: run a live diff against Azure using a .bicepparam file.

Workflow

1) Plan the Architecture

  • Gather comprehensive requirements using the generate_infrastructure_plan MCP tool.
  • This returns:
    • Overview of the infrastructure design
    • Complete list of required resources with types and API versions
    • Resource relationships and dependencies
    • Security and configuration requirements
  • Explain the proposed architecture to the user before proceeding.

2) Generate Resource Configurations

  • For each resource in the plan, use the generate_resource_body MCP tool to get properly configured resource definitions.
  • This returns complete resource body JSON with security best practices applied.
  • Ensure all resources from the plan are accounted for.

3) Create the Bicep Files

  • Generate complete .bicep and .bicepparam files including:
    • Parameters with @description annotations for configurable values
    • Variables for computed values
    • Resources converted from JSON to Bicep syntax
    • Proper resource dependencies using Bicep references
    • Outputs for important resource properties (IDs, endpoints, URLs)
    • Comments documenting configuration decisions
  • Use a main entrypoint (for example main.bicep) and split resources into modules by domain if the infrastructure is complex.
  • Declare parameter values in the .bicepparam file, rather than using default values in the .bicep file parameter declarations.

4) Offline validation

  • After generating files, run get_bicep_file_diagnostics on the .bicep files.
  • Fix compilation errors immediately.
  • Review warnings and fix anything that could affect correctness, deployment behavior, or maintainability.

5) Live validation

  • Run what_if_deployment on the .bicepparam file.
  • Verify the output matches your epxectations.

Acceptance Criteria

  • A coherent set of .bicep files + a .bicepparam file exists for the requested infrastructure.
  • The Bicep compiles cleanly (diagnostics show no errors; warnings are understood/acceptable).
  • All resources from the infrastructure plan are included.
  • Security best practices from the generated resource bodies are applied.

スコア

総合スコア

60/100

リポジトリの品質指標に基づく評価

SKILL.md

SKILL.mdファイルが含まれている

+20
LICENSE

ライセンスが設定されている

+10
説明文

100文字以上の説明がある

0/10
人気

GitHub Stars 100以上

0/15
最近の活動

3ヶ月以内に更新がある

0/10
フォーク

10回以上フォークされている

0/5
Issue管理

オープンIssueが50未満

+5
言語

プログラミング言語が設定されている

+5
タグ

1つ以上のタグが設定されている

0/5

レビュー

💬

レビュー機能は近日公開予定です