Custom LLM Compiler Pipeline Builder

Design and deploy tailored LLM-powered compilation workflows that transform natural language specifications into optimized, production-ready code across multiple target languages and frameworks.

<role>You are an expert LLM Compiler Architect specializing in designing custom compilation pipelines that leverage large language models for code generation, optimization, and multi-target transpilation.</role>

<task>Create a comprehensive custom LLM compiler configuration that translates [input_specification_type] into optimized [target_language] code with [optimization_goals] while maintaining [quality_requirements].</task>

<context>
  <project_overview>[project_name] requires a specialized compilation pipeline that bridges high-level specifications and executable code using LLM-based transformation stages.</project_overview>
  <input_specifications>
    <source_format>[input_specification_type]</source_format>
    <domain_context>[domain_context]</domain_context>
    <complexity_level>[complexity_level]</complexity_level>
  </input_specifications>
  <target_environment>
    <primary_language>[target_language]</primary_language>
    <framework_stack>[framework_stack]</framework_stack>
    <deployment_target>[deployment_target]</deployment_target>
  </target_environment>
  <performance_requirements>
    <optimization_priorities>[optimization_goals]</optimization_priorities>
    <quality_standards>[quality_requirements]</quality_standards>
    <compliance_needs>[compliance_requirements]</compliance_needs>
  </performance_requirements>
</context>

<constraints>
  <constraint>Use only approved LLM models: [approved_models]</constraint>
  <constraint>Maintain deterministic output for identical inputs via [determinism_strategy]</constraint>
  <constraint>Enforce security policies: [security_policies]</constraint>
  <constraint>Keep compilation latency under [max_latency_ms]ms per module</constraint>
  <constraint>Support incremental compilation with [incremental_strategy]</constraint>
  <constraint>Generate comprehensive audit trails for [audit_requirements]</constraint>
</constraints>

<format>
  <output_structure>
    <pipeline_config>
      <stage name="parsing" model="[parser_model]" prompt_template="[parser_prompt]" validation="[parser_validation]"/>
      <stage name="analysis" model="[analyzer_model]" prompt_template="[analyzer_prompt]" validation="[analyzer_validation]"/>
      <stage name="generation" model="[generator_model]" prompt_template="[generator_prompt]" validation="[generator_validation]"/>
      <stage name="optimization" model="[optimizer_model]" prompt_template="[optimizer_prompt]" validation="[optimizer_validation]"/>
      <stage name="verification" model="[verifier_model]" prompt_template="[verifier_prompt]" validation="[verifier_validation]"/>
    </pipeline_config>
    <integration_interfaces>
      <api_specification>[api_spec]</api_specification>
      <cli_commands>[cli_spec]</cli_commands>
      <ide_extension>[ide_spec]</ide_extension>
    </integration_interfaces>
    <monitoring_dashboard>
      <metrics>[key_metrics]</metrics>
      <alerting_rules>[alert_rules]</alerting_rules>
    </monitoring_dashboard>
  </output_structure>
  <deliverable_format>YAML configuration with embedded prompt templates and JSON schema validation</deliverable_format>
</format>

<tone>Technical, precise, and implementation-focused with emphasis on reliability, observability, and developer experience.</tone>

<final_instruction>Generate the complete custom LLM compiler pipeline configuration now, replacing all placeholders with your specific project requirements.</final_instruction>