openapi-toolkit-open-responses-dart

OpenAPI Toolkit (open_responses_dart)

Uses shared scripts from openapi-toolkit with open_responses_dart-specific configuration.

Prerequisites

• Working directory: Repository root
• Python 3
• No API key required to fetch spec (it's public)
• For integration tests: OPENAI_API_KEY environment variable

Spec Registry

Workflow

1. Fetch Latest Spec

python3 .claude/shared/openapi-toolkit/scripts/fetch_spec.py \
  --config-dir .claude/skills/openapi-toolkit-open-responses-dart/config

Output: /tmp/openapi-toolkit-open-responses-dart/latest-main.json

2. Analyze Changes

python3 .claude/shared/openapi-toolkit/scripts/analyze_changes.py \
  --config-dir .claude/skills/openapi-toolkit-open-responses-dart/config \
  packages/open_responses_dart/specs/openapi.json \
  /tmp/openapi-toolkit-open-responses-dart/latest-main.json \
  --format all \
  --changelog-out /tmp/openapi-toolkit-open-responses-dart/changelog.md \
  --plan-out /tmp/openapi-toolkit-open-responses-dart/plan.md

Generates:

• changelog.md - Human-readable change summary
• plan.md - Prioritized implementation plan (P0-P4)

3. Implement Changes

Before implementing, read references/implementation-patterns.md for:

• Model class structure and conventions
• Sealed class patterns for polymorphic types
• Streaming event handling
• JSON serialization patterns

Use templates from ../../shared/openapi-toolkit/assets/.

3.5 Update Documentation (MANDATORY)

Before running the review checklist, update all documentation:

1. README.md - Add/update new features
2. example/ - Create/update example files
3. CHANGELOG.md - Add entry for new features/changes

4. Review & Validate (MANDATORY)

Perform the multi-pass review documented in references/REVIEW_CHECKLIST.md:

# Pass 2: Barrel file verification
python3 .claude/shared/openapi-toolkit/scripts/verify_exports.py \
  --config-dir .claude/skills/openapi-toolkit-open-responses-dart/config

# Pass 3: Documentation completeness
python3 .claude/shared/openapi-toolkit/scripts/verify_readme.py \
  --config-dir .claude/skills/openapi-toolkit-open-responses-dart/config
python3 .claude/shared/openapi-toolkit/scripts/verify_examples.py \
  --config-dir .claude/skills/openapi-toolkit-open-responses-dart/config

# Pass 4: Property-level verification
python3 .claude/shared/openapi-toolkit/scripts/verify_model_properties.py \
  --config-dir .claude/skills/openapi-toolkit-open-responses-dart/config

# Dart quality checks
cd packages/open_responses_dart && dart analyze --fatal-infos && dart format --set-exit-if-changed . && mcp__dart__run_tests()

5. Finalize

# Copy fetched spec to persisted location
cp /tmp/openapi-toolkit-open-responses-dart/latest-main.json packages/open_responses_dart/specs/openapi.json

Package-Specific References

• Package Guide - Package structure, naming conventions
• Implementation Patterns - Model conventions, serialization patterns
• Review Checklist - Multi-pass validation process

External References

Specification

• OpenResponses Home
• OpenResponses Specification
• OpenResponses Changelog
• OpenResponses OpenAPI Spec

Provider Documentation

• OpenAI Responses API
• Vercel AI Gateway
• vLLM OpenResponses
• vLLM MCP Tools
• OpenRouter Responses
• OpenRouter Reasoning
• OpenRouter Tool Calling
• HuggingFace Responses
• Ollama Compatibility
• LM Studio

Troubleshooting

• Network errors: Check connectivity; retry after a few seconds
• No changes detected: Summary shows all zeros; no action needed
• Integration test failures: Set OPENAI_API_KEY environment variable