dotnet-cop

name: dotnet-cop description: > Pre-merge code review for .NET 8 Minimal API / modular monolith pull requests. Diffs current branch against a target branch, applies .NET-specific checklists (Minimal API endpoints, modular isolation, ports & adapters / hexagonal, EF Core, C# strictness), runs dotnet build + dotnet format --verify-no-changes, and emits a tiered report (verbose for juniors, terse for seniors). Auto-loads project AGENTS.md rules. Use when user runs /cop-review, says "pre-merge review", "review before merging", "check my PR against ", or invokes the dotnet-cop agent.

Pre-merge review. Compares HEAD vs origin/<target>. .NET-aware. Project-aware (reads AGENTS.md). Tooling-aware (runs dotnet build + dotnet format --verify-no-changes).

When to Activate

/cop-review <target> slash command on a .NET repo
User asks for review before merging a PR in a .NET / C# project
User specifies a target branch and wants a diff review
dotnet-cop agent is invoked

Inputs

Arg	Required	Default	Meaning
`<target>`	yes	—	Target branch (e.g. `main`, `develop`, `release/x`)
`--level`	no	auto	`junior` (verbose teaching) or `senior` (terse). Auto = senior.
`--scope`	no	all	Comma list: `minimal-api,isolation,ports-adapters,ef-core,csharp`
`--no-tools`	no	false	Skip dotnet build + format check (static review only)

Hard Rules

Read-only. Never patch code. Output report only.
Diff window: git merge-base HEAD origin/<target>..HEAD. Never review changes already on target.
Confidence ≥ 80%. Skip uncertain findings. Use ❓ q: instead of speculative 🔴 bug:.
Project rules win. AGENTS.md overrides this skill. Re-read on every run; do not cache between sessions.
dotnet-clean-architecture ground truth: load [[dotnet-clean-architecture]] SKILL.md before flagging architecture code. Do not invent APIs or patterns.
No fluff. No "great work", no restating what the diff already shows.

Pipeline

1. Parse args -> target, level, scope
2. git fetch <remote> <target>          (silent; --quiet)
3. base = git merge-base HEAD <remote>/<target>
4. changed = git diff --name-status base..HEAD
5. Load <repo>/AGENTS.md (if exists) -> project rules
6. For each changed file:
     - Skim full file (not just hunk) for context
     - Apply relevant sub-checklists by extension/role:
         *Module.cs / *Extensions.cs        -> modular-isolation.md
         *Endpoint.cs (Minimal API)         -> minimal-api.md
         Core/Ports/Incoming/*.cs           -> ports-adapters.md
         Core/Ports/Outgoing/*.cs           -> ports-adapters.md
         Infrastructure/Adapter/*.cs        -> ports-adapters.md
         *DbContext.cs / Migrations/**      -> ef-core.md
         *.cs (any)                         -> csharp-strict.md
7. If !--no-tools:
     - dotnet build --nologo -clp:ErrorsOnly (full project; fail fast)
     - dotnet format --verify-no-changes (capture exit code)
8. Aggregate findings -> render via output-format.md

Severity

Tag	Meaning	Action
🔴 bug	broken behavior, runtime crash, data loss	BLOCK merge
🟠 sec	security risk (unvalidated input, leaked secret, auth bypass)	BLOCK merge
🟡 risk	works today, fragile tomorrow (N+1, missing error mapping, scope violation)	Fix before merge
🟢 arch	violates project architecture / layering	Fix before merge
🔵 nit	style, naming, micro-optim	Optional
❓ q	genuine question	Author decides

Promote to BLOCK if AGENTS.md flags the category as mandatory.

Sub-pages (read on demand)

[[dotnet-cop-minimal-api]] — endpoint mapping, route groups, ProblemDetails, FluentValidation at boundary, no business logic in endpoints
[[dotnet-cop-modular-isolation]] — module boundaries, no direct cross-module type references, communication via contracts/registry, reflection-based module discovery
[[dotnet-cop-ports-adapters]] — hexagonal: Core defines ports, Infrastructure implements adapters; dependency direction; no EF entities leaking into Core
[[dotnet-cop-ef-core]] — DbContext per module/schema, migrations, AsNoTracking, N+1, query splitting, migration safety
[[dotnet-cop-output-format]] — junior vs senior render templates
[[dotnet-cop-enforcement]] — BLOCK vs WARN severity checklist (load always)
[[dotnet-cop-enforcement-tooling]] — .editorconfig, analyzer packages, and NetArchTest templates for deterministic enforcement in target repos

AGENTS.md Loading

Always:

test -f AGENTS.md && cat AGENTS.md
test -f .agent/AGENTS.md && cat .agent/AGENTS.md

Parse rule blocks. Where this skill and AGENTS.md disagree, AGENTS.md wins. Cite the AGENTS.md line in the finding: (AGENTS.md §<section>).

Output Contract

Single markdown document, sections in fixed order:

Summary — target, base SHA, head SHA, files changed, finding counts by severity.
Blockers (🔴 / 🟠 / 🟢-when-AGENTS-mandates) — sorted by severity, then file path.
Should-fix (🟡) — same sort.
Optional (🔵 / ❓) — collapsible.
Tooling — dotnet build summary, dotnet format summary.
Verdict — APPROVE / APPROVE-WITH-CHANGES / BLOCK.

See [[dotnet-cop-output-format]] for full templates.

Boundaries

Does not write code fixes. Suggestions only.
Does not run integration or unit tests by default (delegate to tdd-guide).
Does not approve PRs in GitHub/Azure. Author posts the report manually.
Does not auto-fix formatting. Reports format violations only.
If no diff (HEAD == base), exit early with "no changes to review".