Cross-references C# Web API controllers and DTOs against TypeScript/JavaScript consumers to detect contract drift and breaking changes.
- GitHub stars
- 41,038
- Files
- 1
- Last commit
- Nov 11, 2026
- License
- MIT
- Allowed tools
- —
What does the api-breaking-change-detector skill do?
api-breaking-change-detector is an agent skill. Cross-references C# Web API controllers and DTOs against TypeScript/JavaScript consumers to detect contract drift and breaking changes.
Summary generated from the skill's SKILL.md.
About the api-breaking-change-detector skill
Detecting contract drift between C# Web API controllers and TypeScript or JavaScript consumers is what this skill enables. It helps developers identify when backend changes, such as renamed JSON keys or new required parameters, will break client applications like React, Angular, or Vue. It also identifies when frontend code sends fields that the backend no longer processes, helping to catch silent runtime bugs before they reach production.
An agent should load this tool when a user needs to audit a DTO or controller change against its consumers before merging code. Unlike tools that compare OpenAPI specification files, this skill works directly against the actual source code. It analyzes Program.cs for JSON naming policies, inspects C# 11 required modifiers, and uses normalized URL matching to find hand-written HTTP calls using Fetch or Axios. The process produces a detailed audit report instead of modifying any files.
What an agent can do with it
- 1Audit DTO changes for breaking impactsCheck if modifying a C# Data Transfer Object will break existing frontend interfaces.
- 2Verify API contract synchronizationEnsure that the backend controllers and the TypeScript client models remain in sync.
- 3Perform pre-release endpoint auditsAudit the entire API surface for potential breaking changes before a production release.
Written from the skill's SKILL.md. Every use case is checked against a quote from it.
How to install the api-breaking-change-detector skill
Each command copies the skill folder into the client's skills directory. Restart the client or start a new session to load it.Install the api-breaking-change-detector skill in Claude Code
~/.claude/skills/api-breaking-change-detector/SKILL.md
$ npx degit github/awesome-copilot/skills/api-breaking-change-detector ~/.claude/skills/api-breaking-change-detectorInstall the api-breaking-change-detector skill in Claude Code (project)
.claude/skills/api-breaking-change-detector/SKILL.md
$ npx degit github/awesome-copilot/skills/api-breaking-change-detector .claude/skills/api-breaking-change-detectorInstall the api-breaking-change-detector skill in Codex
~/.agents/skills/api-breaking-change-detector/SKILL.md
$ npx degit github/awesome-copilot/skills/api-breaking-change-detector ~/.agents/skills/api-breaking-change-detectorInstall the api-breaking-change-detector skill in Gemini CLI
~/.gemini/skills/api-breaking-change-detector/SKILL.md
$ npx degit github/awesome-copilot/skills/api-breaking-change-detector ~/.gemini/skills/api-breaking-change-detectorInstall the api-breaking-change-detector skill in Claude.ai
npx degit github/awesome-copilot/skills/api-breaking-change-detector api-breaking-change-detector && zip -r api-breaking-change-detector.zip api-breaking-change-detector # claude.ai → Customize → Skills → + → Create skill → Upload a skill → api-breaking-change-detector.zip
Automated checks
Ran Nov 15, 2026Full report →Passed our automated checks. That is not a security audit: read the scripts before you let an agent run them. How checks work
- Source repository linked and reachablegithub/awesome-copilot/skills/api-breaking-change-detector is public and reachablegithub.com/github/awesome-copilot/tree/HEAD/skills/api-break ↗Pass
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit 2026-11-11github.com/github/awesome-copilot/commits ↗Pass
- No download-and-run or credential theft in scriptsNo scripts in the skill folderPass
- No hidden instructions in SKILL.mdNo override phrases, safeguard switches, credential requests or hidden charactersgithub.com/github/awesome-copilot/blob/HEAD/skills/api-break ↗Pass
- Pre-approved tools are scopedNo pre-approved tools; the agent asks before using onegithub.com/github/awesome-copilot/blob/HEAD/skills/api-break ↗Pass
- SKILL.md follows the Agent Skills formatValid name and description; says when to use it; 73 linesgithub.com/github/awesome-copilot/blob/HEAD/skills/api-break ↗Pass
SKILL.md
View on GitHubAPI Breaking Change Detector
You are cross-referencing a C# Web API's actual contract (controllers, DTOs, route definitions) against its TypeScript/JavaScript consumers to find contract drift — in both directions — before it reaches production.
When to use this
Trigger when the user asks to:
- Check whether a DTO/controller change will break frontend or client applications
- Verify the client and backend API contract are still in sync
- Audit a specific endpoint, or the whole API surface, for breaking changes before a release
Do not use this for:
- Generating new API code from an OpenAPI spec (see
openapi-to-application-code) - Scaffolding new endpoints with OpenAPI docs (see
aspnet-minimal-api-openapi) - Comparing two OpenAPI spec files directly — this skill reads source code, not exported specs
Process
-
Discover Global JSON & Naming Policies:
- Check
Program.csorStartup.csfor active JSON options (e.g.JsonNamingPolicy.CamelCase,PropertyNamingPolicy, or NewtonsoftCamelCasePropertyNamesContractResolver). - Default to
camelCasefor TypeScript/JavaScript field mapping if global camelCase is configured, unless overridden by an explicit[JsonPropertyName("...")]attribute on the C# property. - Ignore C# properties annotated with
[JsonIgnore].
- Check
-
Identify the C# contract surface. For each Controller action in scope:
- Route: Base
[Route("...")]+ action[HttpGet("...")]/[HttpPost("...")]. Normalize route parameters (e.g.{id:int}or{id:guid}$\rightarrow${id}). - Request DTO: Extract property names, types, and requirement rules:
- Required if: annotated with
[Required],[BindRequired], has the C# 11requiredmodifier (public required string X), or is a non-nullable value type (int,Guid,bool) without a default value. - Optional if: nullable (
string?,int?), or has a default initializer.
- Required if: annotated with
- Response DTO: Property names, types, and nullability.
- Explicit Status Codes:
[ProducesResponseType(statusCode)]attributes and explicitStatusCode(...)return paths.
- Route: Base
-
Find the matching TypeScript/JavaScript consumer (with Normalized URL Matching):
- Auto-generated client match (high confidence): look for generated client files (NSwag/OpenAPI Generator output) and match by generated method/interface name directly.
- Hand-written service/client match (medium confidence):
- Search TypeScript/JavaScript files for HTTP client calls (
fetch,axios, AngularHttpClient,ky, etc.) whose normalized URL pattern matches the controller's route. - Normalize template strings and concatenations (e.g.,
${this.apiUrl}/users/${id}orbaseUrl + '/users/' + userId$\rightarrow$/users/{id}). - Match normalized routes against C# routes regardless of variable naming in TypeScript/JS.
- Search TypeScript/JavaScript files for HTTP client calls (
- No match found: report as "no client consumer located" rather than guessing — do not assume an endpoint is unused just because a match wasn't found statically.
- Label every finding with which of these three methods was used to locate it.
-
Compare backend → frontend/client (breaks the client):
- A DTO property renamed or removed that the TypeScript/JS interface or object model still expects
- A new required request field the client never sends
- A response status code the client doesn't handle (e.g. controller now returns 409 Conflict, but client error handler only handles 400/500)
- A response field's type changed (e.g.
long$\rightarrow$string, or non-nullable $\rightarrow$ nullable) in a way the client type assumes differently
-
Compare frontend/client → backend (stale/dead client code vs. silent bugs):
- Harmless dead field: Client sends a payload property the backend ignores without error.
- Silently broken bug (High Severity): Client logic reads a response property that the backend no longer returns (resulting in
undefinedat runtime and potential application failures).
-
Produce the report (see Output Format). This skill does not modify code.
Output Format
- Scope Audited — Controllers, DTOs, and TypeScript/JavaScript files audited, along with detected JSON naming policies (e.g.
camelCaseenabled viaProgram.cs). - Backend → Client Breaks — Grouped by endpoint: what changed, match method used (Auto-generated / Normalized Route Match), exact impact on the client, and severity (Compilation Error vs. Silent Runtime Failure).
- Client → Backend Drift — Stale fields sent or expected, explicitly distinguishing harmless dead fields from silently broken client UI logic.
- No Consumer Found — Unmatched backend DTOs/endpoints requiring manual confirmation.
- Match Confidence Summary — Breakdown of findings derived from auto-generated clients vs. normalized hand-written routes vs. unmatched routes.
Guidelines
- URL Normalization: Always strip query parameters (
?status=active) and normalize path parameters (${id}/:id/{id}) before comparing routes. - Naming Policies: Never assume a C# property name matches a TypeScript/JS property verbatim without checking for
[JsonPropertyName("...")]or globalJsonNamingPolicy.CamelCasesettings. - Modern C# Nuances: Check for C# 11
requiredkeyword and#nullable enableannotations (string?vsstring) when assessing required properties. - Framework Agnostic: Apply contract matching across any TypeScript or JavaScript client (Fetch, Axios, Angular, React, Vue, Svelte, Node.js).
- Never Fabricate: If no matching client service or DTO is found, report "No consumer located via static search" — never guess a pairing based purely on loose class names.
- Reporting Only: Do not modify code; output a scannable, actionable audit report.
Files in this skill
1api-breaking-change-detector skill questions
How do I install the api-breaking-change-detector skill in Claude Code?
Run npx degit github/awesome-copilot/skills/api-breaking-change-detector ~/.claude/skills/api-breaking-change-detector. Claude Code loads skills from ~/.claude/skills (all projects) or .claude/skills (one project) and uses a skill when its description matches the task.
Which clients can use the api-breaking-change-detector skill?
Any client that reads the Agent Skills format: Claude Code, Codex, Gemini CLI and claude.ai (upload as a ZIP). Scripts in the skill run only where the client allows code execution.
Does this tool use OpenAPI spec files for comparison?
No, this skill works directly against source code rather than using exported OpenAPI spec files.
What kind of frontend clients can be audited?
It supports various TypeScript/JavaScript consumers including React, Angular, Vue, Svelte, Node.js, and clients using Fetch, Axios, or NSwag.
How does the tool handle C# naming policies?
It checks Program.cs or Startup.cs for JSON naming policies like camelCase and respects explicit [JsonPropertyName] attributes.
Will this tool modify my source code?
No, the tool is designed to produce an audit report only and does not perform any code modifications.