Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generate a TypeSpec API plugin with REST operations, authentication, and Adaptive Cards for Microsoft 365 Copilot
.claude/skills/typespec-create-api-plugin/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flash | 100% | 9 |
| gemini-3.1-pro-previewlowest | 75% | 8 |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | — | — |
| case-07 | ✗→✓ | ▲ Improved | — | — |
| case-21 | ✗→✓ | ▲ Improved | — | — |
| case-18 | ✗→✓ | ▲ Improved | — | — |
| case-14 | ✗→✓ | ▲ Improved | — | — |
Create a complete TypeSpec API plugin for Microsoft 365 Copilot that integrates with external REST APIs.
Generate TypeSpec files with:
typescriptimport "@typespec/http"; import "@typespec/openapi3"; import "@microsoft/typespec-m365-copilot"; import "./actions.tsp"; using TypeSpec.Http; using TypeSpec.M365.Copilot.Agents; using TypeSpec.M365.Copilot.Actions; @agent({ name: "[Agent Name]", description: "[Description]" }) @instructions(""" [Instructions for using the API operations] """) namespace [AgentName] { // Reference operations from actions.tsp op operation1 is [APINamespace].operationName; }
typescriptimport "@typespec/http"; import "@microsoft/typespec-m365-copilot"; using TypeSpec.Http; using TypeSpec.M365.Copilot.Actions; @service @actions(#{ nameForHuman: "[API Display Name]", descriptionForModel: "[Model description]", descriptionForHuman: "[User description]" }) @server("[API_BASE_URL]", "[API Name]") @useAuth([AuthType]) // Optional namespace [APINamespace] { @route("[/path]") @get @action op operationName( @path param1: string, @query param2?: string ): ResponseModel; model ResponseModel { // Response structure } }
Choose based on API requirements:
typescript // No @useAuth decorator needed
typescript @useAuth(ApiKeyAuth<ApiKeyLocation.header, "X-API-Key">)
typescript @useAuth(OAuth2Auth<[{ type: OAuth2FlowType.authorizationCode; authorizationUrl: "https://oauth.example.com/authorize"; tokenUrl: "https://oauth.example.com/token"; refreshUrl: "https://oauth.example.com/token"; scopes: ["read", "write"]; }]>)
typescript @useAuth(Auth)
@authReferenceId("registration-id-here") model Auth is ApiKeyAuth<ApiKeyLocation.header, "X-API-Key">
typescript@capabilities(#{ confirmation: #{ type: "AdaptiveCard", title: "Confirm Action", body: """ Are you sure you want to perform this action? * **Parameter**: {{ function.parameters.paramName }} """ } })
typescript@card(#{ dataPath: "$.items", title: "$.title", url: "$.link", file: "cards/card.json" })
typescript@reasoning(""" Consider user's context when calling this operation. Prioritize recent items over older ones. """) @responding(""" Present results in a clear table format with columns: ID, Title, Status. Include a summary count at the end. """)
Ask the user:
Then generate:
main.tsp with agent definitionactions.tsp with API operations and modelscards/card.json if Adaptive Cards are needed| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-19 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-16 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +77 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
Other measured skills in the registry, with their headline benchmark lift.