Skip to content

Commit 0330580

Browse files
ZangesZanges
authored
Add code quality automation: analyzers, Dependabot auto-merge, PR labeler, multi-agent Claude review (#9)
2 parents 5eb5a0b + c6a61bc commit 0330580

113 files changed

Lines changed: 5226 additions & 467 deletions

File tree

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

‎.claude/skills/review-pr/SKILL.md‎

Lines changed: 178 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,178 @@
1+
---
2+
name: review-pr
3+
description: Run the Mouse2Joy multi-angle review locally on a PR or the current branch. Spawns 8 review subagents (security, correctness, regression, ux, stability, convention-compliance, performance, code-reuse) using the same reviewer prompts as the CI pipeline. Reports findings in the conversation instead of posting to GitHub. Use when the user wants to review a PR before opening it, double-check a branch, or get a second opinion on changes without waiting for CI.
4+
---
5+
6+
# review-pr
7+
8+
This is the **interactive / local** counterpart to the CI workflow at
9+
[`.github/workflows/claude-review.yml`](../../../.github/workflows/claude-review.yml).
10+
Same 8 reviewer angles, same prompts, but it runs from a Claude Code session
11+
and reports findings as text in the conversation instead of posting a GitHub
12+
review.
13+
14+
## When to use
15+
16+
Trigger this skill when the user asks for things like:
17+
18+
- "Review this PR" / "review PR #12" / "give me a review of branch X"
19+
- "Check my branch before I open the PR"
20+
- "What would the CI reviewer flag on this?"
21+
- "Run the multi-agent review locally"
22+
23+
Do **not** trigger if the user just wants a quick code-quality scan of one
24+
file or a specific question about one change — those don't need 8 angles.
25+
26+
## What it does
27+
28+
1. Resolves the target: PR number (via `gh pr view <n>`), explicit branch
29+
name, or "the current branch" by default.
30+
2. Loads the 8 reviewer system prompts from
31+
[`.github/claude-review/reviewers/`](../../../.github/claude-review/reviewers/).
32+
These files are the **single source of truth** — they are also used by
33+
the CI workflow.
34+
3. Lists the changed files in the diff.
35+
4. Runs each reviewer as a `Task` subagent in turn (sequentially, not in
36+
parallel — the local skill optimizes for transparency in the conversation
37+
over throughput).
38+
5. Aggregates findings into a single response in the conversation, with one
39+
section per angle.
40+
41+
## How to run it
42+
43+
Follow these steps in order. The user is in this Claude Code session; you
44+
are the orchestrator.
45+
46+
### Step 1: Resolve the target
47+
48+
Look at the user's request:
49+
50+
- If they named a PR number (e.g. "review PR #8"): use that.
51+
- If they named a branch: use `git diff <base>...<branch>` against
52+
`origin/main`.
53+
- Otherwise: default to the current branch vs `origin/main`. Run
54+
`git rev-parse --abbrev-ref HEAD` and check that it isn't `main`. If it
55+
is, ask the user which branch / PR they meant.
56+
57+
### Step 2: Identify changed files
58+
59+
Run one of:
60+
61+
- For a PR: `gh pr view <n> --json files --jq '.files[].path'`
62+
- For a branch: `git diff --name-only origin/main...<branch>`
63+
64+
Hold this list — you'll pass it to every reviewer.
65+
66+
If the list is empty, say so and stop.
67+
68+
If the list is *only* docs (`ai-docs/**`, `**/*.md`, `LICENSE`) or *only*
69+
dependency files (`Directory.Packages.props`, `global.json`, csproj package
70+
references), tell the user "this matches the CI gate's skip rules and
71+
wouldn't be reviewed automatically — running anyway since you asked
72+
explicitly" and continue.
73+
74+
### Step 3: Load reviewer prompts
75+
76+
Read all 8 files from `.github/claude-review/reviewers/`:
77+
78+
- `security.md`
79+
- `correctness.md`
80+
- `regression.md`
81+
- `ux.md`
82+
- `stability.md`
83+
- `convention-compliance.md`
84+
- `performance.md`
85+
- `code-reuse.md`
86+
87+
You'll inject each one as the `Task` subagent's instructions.
88+
89+
### Step 4: Run reviewers sequentially
90+
91+
For each of the 8 reviewer prompts, invoke the `Task` tool with:
92+
93+
- `subagent_type: "general-purpose"`
94+
- `description`: short (e.g. "Security review of PR #8")
95+
- `prompt`: the full text of the reviewer prompt file, followed by:
96+
97+
```
98+
---
99+
100+
## This review
101+
102+
Target: <PR #N | branch <name>>
103+
Base: origin/main
104+
Head: <head ref>
105+
106+
Changed files:
107+
<bulleted list of paths>
108+
109+
Use Read / Grep / Glob / Bash (for `git diff` and `gh pr diff`) to
110+
investigate. Don't dump everything — pull only what you actually need
111+
for your angle. Output the structured JSON your instructions require.
112+
```
113+
114+
After each subagent returns, summarize its findings into the conversation
115+
under a short heading like `## Security (N findings)` so the user sees
116+
progress, then move on. Don't try to render the JSON literally — turn it
117+
into readable markdown:
118+
119+
- Summary as a sentence.
120+
- Critical findings as bullets with `path:L<line>` references.
121+
- Suggestions / nits collapsed under a "more" or omitted if low value (the
122+
user is reading in real-time; respect their attention).
123+
124+
### Step 5: Final wrap-up
125+
126+
After all 8 finish, post one short consolidated summary:
127+
128+
- "**Critical**: N findings" with a one-line description each.
129+
- "**Suggestions worth considering**: N" — list 3–5 of the most useful.
130+
- Token usage if you tracked it (optional).
131+
132+
Offer to dive deeper into any specific finding the user wants to discuss.
133+
134+
## Important constraints
135+
136+
- **Do NOT post to GitHub.** This skill is local-only. The CI workflow
137+
handles GitHub reviews. If the user wants the same review posted as a
138+
GitHub review, point them at the CI pipeline (push the branch, open
139+
the PR, CI runs the review automatically).
140+
- **Sequential, not parallel.** The CI script fan-outs 8 agents in
141+
parallel; this skill runs them one at a time so the user sees progress
142+
and can interrupt. If they want speed, the CI workflow exists.
143+
- **Reviewer prompts are the contract.** Don't paraphrase or summarize
144+
them when injecting into a subagent — pass the file content verbatim.
145+
This is what keeps CI review and local review in sync.
146+
- **Conventions referenced by the prompts** (`CLAUDE.md`, the
147+
"Things that are easy to get wrong" section, `ai-docs/implementations/`)
148+
are read by the subagents themselves via tool calls. You don't need to
149+
pre-load them; the agent will reach for them when relevant.
150+
151+
## What this skill does NOT do
152+
153+
- Open PRs / push branches / merge anything.
154+
- Modify files (it's a review, not a fix).
155+
- Run the .NET build or tests (CI does that; this skill is post-CI in
156+
spirit, even if there's no formal gate).
157+
- Replace [`/review`](https://docs.claude.com/claude-code) — that's the
158+
generic Claude Code review command. This skill is the Mouse2Joy-specific
159+
8-angle methodology.
160+
161+
## Example invocations
162+
163+
User: "Review my current branch."
164+
165+
You:
166+
1. `git rev-parse --abbrev-ref HEAD` → `feature/widget-rotation`.
167+
2. `git diff --name-only origin/main...feature/widget-rotation` → 4 files.
168+
3. Read the 8 reviewer prompts.
169+
4. Run each via `Task` in turn, posting per-angle summaries as you go.
170+
5. Final consolidated summary.
171+
172+
---
173+
174+
User: "Run the multi-agent review on PR #12."
175+
176+
You:
177+
1. `gh pr view 12 --json files,headRefName --jq ...`
178+
2. Same as above, but with PR-specific context.

‎.editorconfig‎

Lines changed: 90 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,90 @@
1+
# EditorConfig: https://editorconfig.org
2+
#
3+
# Pairs with <AnalysisLevel>latest-recommended</AnalysisLevel> +
4+
# <EnforceCodeStyleInBuild>true</EnforceCodeStyleInBuild> in Directory.Build.props.
5+
#
6+
# Philosophy: keep correctness/perf rules on, silence stylistic rules that
7+
# would generate noise on a working WPF + native-interop codebase without
8+
# improving quality. Severities here override analyzer defaults at build
9+
# time. With TreatWarningsAsErrors=true, anything left at "warning" becomes
10+
# a build failure -- choose between "warning" (must fix) and "suggestion"
11+
# (IDE-only hint) carefully.
12+
13+
root = true
14+
15+
[*]
16+
charset = utf-8
17+
end_of_line = crlf
18+
insert_final_newline = true
19+
trim_trailing_whitespace = true
20+
indent_style = space
21+
indent_size = 4
22+
23+
[*.{yml,yaml,json,md}]
24+
indent_size = 2
25+
26+
[*.{csproj,props,targets,xaml}]
27+
indent_size = 2
28+
29+
[*.cs]
30+
# ---- C# language conventions ----
31+
# Use the .NET defaults; keep these as suggestions so they appear in the IDE
32+
# but never break the build.
33+
dotnet_style_qualification_for_field = false:suggestion
34+
dotnet_style_qualification_for_property = false:suggestion
35+
dotnet_style_qualification_for_method = false:suggestion
36+
dotnet_style_qualification_for_event = false:suggestion
37+
csharp_style_var_when_type_is_apparent = true:suggestion
38+
csharp_style_var_elsewhere = false:suggestion
39+
40+
# ---- IDE rule severities (style) ----
41+
# Things we genuinely want enforced.
42+
dotnet_diagnostic.IDE0005.severity = warning # remove unnecessary usings
43+
dotnet_diagnostic.IDE0011.severity = warning # add braces (we already use them)
44+
dotnet_diagnostic.IDE0044.severity = warning # add readonly modifier
45+
dotnet_diagnostic.IDE0051.severity = warning # remove unused private members
46+
dotnet_diagnostic.IDE0052.severity = warning # remove unread private members
47+
48+
# Things we explicitly do NOT want as errors -- stylistic or churny.
49+
dotnet_diagnostic.IDE0008.severity = silent # use explicit type (taste)
50+
dotnet_diagnostic.IDE0058.severity = silent # expression value never used (fluent assertions)
51+
dotnet_diagnostic.IDE0066.severity = suggestion # use switch expression
52+
dotnet_diagnostic.IDE0072.severity = suggestion # add missing cases to switch expression
53+
dotnet_diagnostic.IDE0290.severity = silent # use primary constructor (churny for existing code)
54+
55+
# ---- Code-quality (CA) rule severities ----
56+
# Keep on as warnings (will be errors due to TreatWarningsAsErrors=true).
57+
# Explicitly silence the ones that don't fit a desktop / native-interop app.
58+
59+
dotnet_diagnostic.CA1062.severity = silent # validate args (we use nullable refs at boundaries)
60+
dotnet_diagnostic.CA1805.severity = silent # member initialized to default value (we use it to document persisted defaults on records)
61+
dotnet_diagnostic.CA1303.severity = silent # do not pass literals as localized parameters (no l10n)
62+
dotnet_diagnostic.CA1304.severity = silent # specify CultureInfo (desktop app, system culture is fine)
63+
dotnet_diagnostic.CA1305.severity = silent # specify IFormatProvider (ditto)
64+
dotnet_diagnostic.CA1307.severity = silent # specify StringComparison for clarity (ordinal everywhere by default)
65+
dotnet_diagnostic.CA1310.severity = warning # specify StringComparison for correctness
66+
dotnet_diagnostic.CA1311.severity = silent # specify culture for ToUpper/ToLower
67+
dotnet_diagnostic.CA1716.severity = silent # identifiers should not match keywords
68+
dotnet_diagnostic.CA1724.severity = silent # type names should not match namespaces
69+
dotnet_diagnostic.CA1812.severity = silent # internal class never instantiated (DI containers / WPF)
70+
dotnet_diagnostic.CA1822.severity = suggestion # member can be static (often false-positive for view-models)
71+
dotnet_diagnostic.CA1848.severity = silent # use LoggerMessage delegates (Serilog templates already cached)
72+
dotnet_diagnostic.CA1851.severity = warning # multiple enumerations of IEnumerable
73+
dotnet_diagnostic.CA1860.severity = warning # prefer Count == 0 over !Any()
74+
dotnet_diagnostic.CA2007.severity = silent # ConfigureAwait(false) (WPF app context)
75+
dotnet_diagnostic.CA1051.severity = silent # public readonly fields on hot-path structs (Signal, etc.) are deliberate
76+
77+
# WPF + Win32 P/Invoke realities.
78+
dotnet_diagnostic.CA1031.severity = suggestion # catch general exception types (host code legitimately does this)
79+
dotnet_diagnostic.CA1416.severity = silent # platform compatibility (whole app is win-only)
80+
dotnet_diagnostic.CA1707.severity = silent # underscore in identifiers (Win32 constants / test names)
81+
dotnet_diagnostic.CA1711.severity = silent # identifier suffix rules
82+
dotnet_diagnostic.CA2101.severity = silent # specify marshaling for P/Invoke (LPWStr default is fine)
83+
84+
# Tests: relax a few more things that don't add value in a test project.
85+
[**/*.Tests/**.cs]
86+
dotnet_diagnostic.CA1707.severity = silent # MethodName_With_Underscores is the test convention
87+
dotnet_diagnostic.CA1861.severity = silent # avoid constant arrays as args (test data)
88+
dotnet_diagnostic.CA2007.severity = silent
89+
dotnet_diagnostic.CA1859.severity = silent # return concrete type for perf (tests intentionally use abstractions)
90+
dotnet_diagnostic.IDE0051.severity = silent # unused private member (test helpers paired with peer fixtures)

0 commit comments

Comments
 (0)