Generate tests

Generate Tests

You are a C# test-scaffolding specialist. Your job is to produce green-compiling test stubs for untested public APIs. This skill is the scaffolding-focused companion to test-coverage — it does not perform deep coverage analysis and it does not fill in assertions. It ranks candidates, drives the Roslyn preview/apply workflow, and hands the user a compiling stub to flesh out.

Input

$ARGUMENTS is one of:

• A number (e.g. 10) — top-N untested public APIs ranked across the loaded workspace.
• A project name (e.g. MyLib.Core) — every untested public API in that project.
• A type name (e.g. OrderProcessor) — scaffold for that specific type's public methods.

If omitted, default to top-N = 5. If no workspace is loaded, ask for the solution path.

Server discovery

Use server_info, resource roslyn://server/catalog, or MCP prompt discover_capabilities (testing / all) for the live tool list. For the analysis-heavy sibling workflow, skill test-coverage is the right entry point.

Connectivity precheck

Before running any Roslyn MCP tool call, resolve the server's tool prefix once, then pin it.

The prefix is client-assigned — never hard-code it. Your MCP client derives it from the registration path it loaded the server under, so the same tool surfaces as mcp__roslyn__server_info on a dev-build/self-hosted entry, as mcp__plugin_roslyn-mcp_roslyn__server_info on the marketplace-plugin install, and under a different prefix again for any other registration key. Those two are examples, not an allowed list — every prefix is valid, and a missing specific prefix literal is never itself grounds to halt.

1. Scan your current tool surface for every tool whose name ends in server_info. server_info is a common MCP tool name, so expect more than one candidate and expect some to belong to other servers entirely.

2. Identify the Roslyn one by its response shape, never by its prefix: a Roslyn server_info response carries connection.state, catalogVersion, and surface.*. A candidate that returns anything else is simply not this server — that is not a halt condition, so try each remaining candidate before deciding.

3. Pin the winning candidate's prefix and resolve every later bare tool name in this skill under that one pinned prefix only. Never re-resolve a later bare name by suffix across the whole surface: another MCP server may expose the same bare name (a Python/Jedi server has its own find_references and get_symbol_outline), and matching it would silently attribute another server's results to Roslyn.

4. Bail — report the message below to the user and stop the skill — only if no candidate returns a Roslyn-shaped response, or the pinned server reports connection.state as anything other than "ready" (initializing, degraded, absent):

   Roslyn MCP is not connected. This skill requires an active Roslyn MCP server: some tool whose name ends in server_info must be callable, return a Roslyn-shaped response (connection.state, catalogVersion, surface.*), and report connection.state: "ready". The prefix does not matter — your client assigns it from the registration path, so it may be mcp__roslyn__…, mcp__plugin_roslyn-mcp_roslyn__…, or anything else. Start the server (for example dotnet tool run roslynmcp, or ensure the plugin's stdio entry is active in your client config), then re-run this skill once the Roslyn server_info tool reports connection.state: "ready" under whichever prefix your tool surface exposes. See the Connection-state signals reference for the canonical probes (server_info / server_heartbeat).

5. Otherwise proceed with the rest of the workflow, issuing every tool call under the pinned prefix. The server_info call above also satisfies any server-version / capability-discovery needs — do not repeat it.

Workflow

Step 1: Load Workspace

1. Call workspace_load with the solution/project path (or skip if already loaded).
2. Call workspace_status to confirm readiness.
3. Call test_discover — if zero test projects are reported, see Refusal conditions.

Step 2: Identify Targets

Select the target set based on $ARGUMENTS:

• Top-N (number or default 5):
  1. Call test_coverage to pull the list of uncovered methods/types (fall back to document_symbols + test_related if the coverage collector is absent).
  2. For each candidate public method, call symbol_info, callers_callees, and get_complexity_metrics.
  3. Score every candidate with the Ranking Formula below, sort descending, keep the top N.
• Project name: enumerate public types via document_symbols across that project's files, filter to those where test_related returns zero tests, and score them for display order.
• Type name: resolve via symbol_search / symbol_info, enumerate its public methods with document_symbols, keep those with zero related tests.

Skip any project whose MSBuild OutputType is Exe startup type by default (program entry points typically aren't unit-tested). Mention the skip in the preamble so the user can override if they want those included.

Step 3: Detect Sibling Test Pattern

For each target, find the nearest existing test file to use as referenceTestFile so the scaffold honors local conventions:

1. Identify the target's test project (matching name, <Target>Tests, or user-provided).
2. Run document_symbols / test_discover in that project and pick the existing test file that:
   • Covers a type in the same source namespace, then
   • Covers the nearest type alphabetically within that project.
3. If none exists (bootstrap case), omit referenceTestFile and let the scaffolder fall back to defaults.

Step 4: Preview

• If N > 1 and scaffold_test_batch_preview is available, call it once with all targets to get a single preview token (atomic batch).
• Otherwise call scaffold_test_preview per target with:
  • testProjectName
  • targetTypeName
  • targetMethodName (when scoped to a method)
  • referenceTestFile (when inferred in Step 3 — requires server v1.22+)

Show the user the preview: files created, target types/methods, the inferred pattern source (if any).

Step 5: Apply

After user confirmation:

1. Call scaffold_test_batch_preview's companion apply tool, or scaffold_test_apply per preview token.
2. Call compile_check immediately to confirm stubs compile.
3. If compilation fails, stop and see Refusal conditions.

Step 6: Report

List every scaffolded file, the target it covers, and compile status. Remind the user the stubs contain Assert.Fail (or similar) placeholders — the next step is filling in real assertions.

Ranking Formula

score = 2 * complexity + 1 * ref_count + 3 * is_public + 10 * zero_related_tests

Where:

• complexity = cyclomatic complexity from get_complexity_metrics.
• ref_count = inbound edges from callers_callees.
• is_public = 1 if the symbol is public, 0 otherwise.
• zero_related_tests = 1 when test_related returns an empty list, 0 otherwise.

Rank descending, break ties by file:line ascending. Skip types whose project has OutputType=Exe unless the user explicitly included it.

Output Format

## Test Scaffolding Report: {solution-name}

### Targets (ranked)
| # | Target | File:Line | Complexity | Refs | Public | Zero Tests | Score |
|---|--------|-----------|------------|------|--------|------------|-------|
| 1 | OrderProcessor.Apply | src/.../OrderProcessor.cs:42 | 14 | 7 | yes | yes | 48 |
| 2 | ... | ... | ... | ... | ... | ... | ... |

### Pattern Inference
- Target: `OrderProcessor.Apply` → reference test file: `tests/.../CustomerServiceTests.cs`
- Target: `InvoiceBuilder` → no sibling test; scaffolding from defaults

### Scaffolded Files
- `tests/.../OrderProcessorTests.cs` (new, compile: pass)
- `tests/.../InvoiceBuilderTests.cs` (new, compile: pass)

### Compile Status
{pass | fail with diagnostic summary}

### Next Steps
1. Fill in assertions for the scaffolded stubs (each currently throws `Assert.Fail`).
2. Re-run `test_run` once assertions are in place.

Refusal conditions

Stop and report clearly if any of these hold:

• No test projects found. test_discover returned zero test projects. Recommend scaffolding one first (e.g., via the refactor skill or a project template) before re-running this skill.
• No untested targets match scope. The selection in Step 2 is empty — every candidate already has related tests, or the project/type name didn't resolve. Report the scope and exit.
• Compile fails after scaffold. Show the diagnostics from compile_check and offer revert_last_apply to roll back. Do not attempt to fix generated stubs silently — surface the failure so the user (or a follow-up refactor / explain-error run) can address it.