Use when your spec changed and you need to regenerate the SDK, or running `speakeasy run`
Installation
$skills install @speakeasy-api/regenerate-sdk
Claude Code
Cursor
Copilot
Codex
Antigravity
Details
Repositoryspeakeasy-api/speakeasy
Pathskills/regenerate-sdk/SKILL.md
Branchmain
Scoped Name@speakeasy-api/regenerate-sdk
Usage
After installing, this skill will be available to your AI coding assistant.
Verify installation:
skills listSkill Instructions
name: regenerate-sdk
description: Use when your spec changed and you need to regenerate the SDK, or running speakeasy run
regenerate-sdk
Use speakeasy run to execute the workflow and regenerate SDKs.
Command
# Run all configured targets
speakeasy run
# Run specific target only
speakeasy run -t <target-name>
# Run with specific source
speakeasy run -s <source-name>
# AI-friendly output mode
speakeasy run --output console
When to Use
- After updating the OpenAPI spec
- After modifying workflow.yaml
- After changing overlays
- To regenerate with latest Speakeasy version
Example Workflow
# .speakeasy/workflow.yaml
workflowVersion: 1.0.0
speakeasyVersion: latest
sources:
my-api:
inputs:
- location: ./openapi.yaml
targets:
typescript-sdk:
target: typescript
source: my-api
output: ./sdk/typescript
AI-Friendly Output
For commands with large outputs, pipe to grep or tail to reduce context:
speakeasy run --output console 2>&1 | tail -50
Troubleshooting
If speakeasy run fails, check:
- Is the OpenAPI spec valid? Run
speakeasy lint openapi -s <spec> - Does the source path exist? Check
inputs.locationin workflow.yaml - Are there blocking validation errors? See
diagnose-generation-failureskill
More by speakeasy-api
View allapply-openapi-overlay
377Use when applying an overlay file to a spec
get-ai-suggestions
377Use when SDK method names are ugly, wanting to improve operation IDs, or asking "how can I improve my spec"
merge-openapi-specs
377Use when combining multiple OpenAPI specs, or have microservices with separate spec files
validate-openapi-spec
377Use when checking if an OpenAPI spec is valid, looking for errors, or running `speakeasy lint`
