@the-bearded-bear/claude-craft 9.1.1 → 9.2.0-next.32ff465
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/Dev/i18n/base/Rust/checklists/new-feature.md +108 -0
- package/Dev/i18n/base/Rust/checklists/pre-commit.md +118 -0
- package/Dev/i18n/base/Rust/checks/all.md +251 -0
- package/Dev/i18n/base/Rust/checks/architecture.md +182 -0
- package/Dev/i18n/base/Rust/checks/code-quality.md +174 -0
- package/Dev/i18n/base/Rust/checks/security.md +184 -0
- package/Dev/i18n/base/Rust/checks/testing.md +180 -0
- package/Dev/i18n/base/Rust/commands/check.md +28 -0
- package/Dev/i18n/base/Rust/rules/00-project-context.md.template +253 -0
- package/Dev/i18n/base/Rust/templates/app-error.md +182 -0
- package/Dev/i18n/base/Rust/templates/handler.md +255 -0
- package/Dev/i18n/base/Rust/templates/integration-test.md +193 -0
- package/Dev/i18n/base/Tauri/checklists/new-feature.md +88 -0
- package/Dev/i18n/base/Tauri/checklists/pre-commit.md +121 -0
- package/Dev/i18n/base/Tauri/checks/all.md +250 -0
- package/Dev/i18n/base/Tauri/checks/architecture.md +165 -0
- package/Dev/i18n/base/Tauri/checks/code-quality.md +157 -0
- package/Dev/i18n/base/Tauri/checks/security.md +186 -0
- package/Dev/i18n/base/Tauri/checks/testing.md +153 -0
- package/Dev/i18n/base/Tauri/commands/check.md +28 -0
- package/Dev/i18n/base/Tauri/rules/00-project-context.md.template +161 -0
- package/Dev/i18n/base/Tauri/rules/02-architecture-tauri.md +226 -0
- package/Dev/i18n/base/Tauri/rules/03-coding-standards.md +140 -0
- package/Dev/i18n/base/Tauri/rules/06-tooling.md +133 -0
- package/Dev/i18n/base/Tauri/rules/07-testing-tauri.md +170 -0
- package/Dev/i18n/base/Tauri/rules/08-quality-tools.md +158 -0
- package/Dev/i18n/base/Tauri/rules/11-security-tauri.md +198 -0
- package/Dev/i18n/base/Tauri/templates/capability.md +106 -0
- package/Dev/i18n/base/Tauri/templates/command.md +211 -0
- package/Dev/i18n/base/Tauri/templates/ipc-mock-test.md +188 -0
- package/Dev/i18n/de/Common/rules/12-context-management.md +3 -0
- package/Dev/i18n/de/Rust/CLAUDE.md.template +94 -0
- package/Dev/i18n/de/Rust/agents/rust-reviewer.md +577 -0
- package/Dev/i18n/de/Rust/checks/all.md +250 -0
- package/Dev/i18n/de/Rust/commands/check.md +28 -0
- package/Dev/i18n/de/Rust/rules/02-architecture-rust.md +387 -0
- package/Dev/i18n/de/Rust/rules/03-coding-standards.md +251 -0
- package/Dev/i18n/de/Rust/rules/06-tooling.md +293 -0
- package/Dev/i18n/de/Rust/rules/07-testing-rust.md +333 -0
- package/Dev/i18n/de/Rust/rules/08-quality-tools.md +253 -0
- package/Dev/i18n/de/Rust/rules/11-security-rust.md +420 -0
- package/Dev/i18n/de/Tauri/CLAUDE.md.template +99 -0
- package/Dev/i18n/de/Tauri/agents/tauri-reviewer.md +452 -0
- package/Dev/i18n/de/Tauri/checks/all.md +252 -0
- package/Dev/i18n/de/Tauri/commands/check.md +28 -0
- package/Dev/i18n/en/Common/rules/12-context-management.md +3 -0
- package/Dev/i18n/en/Rust/CLAUDE.md.template +94 -0
- package/Dev/i18n/en/Rust/agents/rust-reviewer.md +577 -0
- package/Dev/i18n/en/Rust/checks/all.md +249 -0
- package/Dev/i18n/en/Rust/commands/check.md +28 -0
- package/Dev/i18n/en/Rust/rules/02-architecture-rust.md +387 -0
- package/Dev/i18n/en/Rust/rules/03-coding-standards.md +251 -0
- package/Dev/i18n/en/Rust/rules/06-tooling.md +293 -0
- package/Dev/i18n/en/Rust/rules/07-testing-rust.md +333 -0
- package/Dev/i18n/en/Rust/rules/08-quality-tools.md +253 -0
- package/Dev/i18n/en/Rust/rules/11-security-rust.md +420 -0
- package/Dev/i18n/en/Tauri/CLAUDE.md.template +99 -0
- package/Dev/i18n/en/Tauri/agents/tauri-reviewer.md +452 -0
- package/Dev/i18n/en/Tauri/checks/all.md +251 -0
- package/Dev/i18n/en/Tauri/commands/check.md +28 -0
- package/Dev/i18n/es/Common/rules/12-context-management.md +3 -0
- package/Dev/i18n/es/Rust/CLAUDE.md.template +94 -0
- package/Dev/i18n/es/Rust/agents/rust-reviewer.md +577 -0
- package/Dev/i18n/es/Rust/checks/all.md +250 -0
- package/Dev/i18n/es/Rust/commands/check.md +28 -0
- package/Dev/i18n/es/Rust/rules/02-architecture-rust.md +387 -0
- package/Dev/i18n/es/Rust/rules/03-coding-standards.md +251 -0
- package/Dev/i18n/es/Rust/rules/06-tooling.md +293 -0
- package/Dev/i18n/es/Rust/rules/07-testing-rust.md +333 -0
- package/Dev/i18n/es/Rust/rules/08-quality-tools.md +253 -0
- package/Dev/i18n/es/Rust/rules/11-security-rust.md +420 -0
- package/Dev/i18n/es/Tauri/CLAUDE.md.template +99 -0
- package/Dev/i18n/es/Tauri/agents/tauri-reviewer.md +452 -0
- package/Dev/i18n/es/Tauri/checks/all.md +252 -0
- package/Dev/i18n/es/Tauri/commands/check.md +28 -0
- package/Dev/i18n/fr/Common/rules/12-context-management.md +3 -0
- package/Dev/i18n/fr/Rust/CLAUDE.md.template +94 -0
- package/Dev/i18n/fr/Rust/agents/rust-reviewer.md +577 -0
- package/Dev/i18n/fr/Rust/checks/all.md +249 -0
- package/Dev/i18n/fr/Rust/commands/check.md +28 -0
- package/Dev/i18n/fr/Rust/rules/02-architecture-rust.md +387 -0
- package/Dev/i18n/fr/Rust/rules/03-coding-standards.md +251 -0
- package/Dev/i18n/fr/Rust/rules/06-tooling.md +293 -0
- package/Dev/i18n/fr/Rust/rules/07-testing-rust.md +333 -0
- package/Dev/i18n/fr/Rust/rules/08-quality-tools.md +253 -0
- package/Dev/i18n/fr/Rust/rules/11-security-rust.md +420 -0
- package/Dev/i18n/fr/Tauri/CLAUDE.md.template +99 -0
- package/Dev/i18n/fr/Tauri/agents/tauri-reviewer.md +452 -0
- package/Dev/i18n/fr/Tauri/checks/all.md +251 -0
- package/Dev/i18n/fr/Tauri/commands/check.md +28 -0
- package/Dev/i18n/pt/Common/rules/12-context-management.md +3 -0
- package/Dev/i18n/pt/Rust/CLAUDE.md.template +94 -0
- package/Dev/i18n/pt/Rust/agents/rust-reviewer.md +577 -0
- package/Dev/i18n/pt/Rust/checks/all.md +249 -0
- package/Dev/i18n/pt/Rust/commands/check.md +28 -0
- package/Dev/i18n/pt/Rust/rules/02-architecture-rust.md +387 -0
- package/Dev/i18n/pt/Rust/rules/03-coding-standards.md +251 -0
- package/Dev/i18n/pt/Rust/rules/06-tooling.md +293 -0
- package/Dev/i18n/pt/Rust/rules/07-testing-rust.md +333 -0
- package/Dev/i18n/pt/Rust/rules/08-quality-tools.md +253 -0
- package/Dev/i18n/pt/Rust/rules/11-security-rust.md +420 -0
- package/Dev/i18n/pt/Tauri/CLAUDE.md.template +99 -0
- package/Dev/i18n/pt/Tauri/agents/tauri-reviewer.md +452 -0
- package/Dev/i18n/pt/Tauri/checks/all.md +251 -0
- package/Dev/i18n/pt/Tauri/commands/check.md +28 -0
- package/Dev/scripts/check-config.sh +1 -1
- package/Dev/scripts/install-from-config.sh +6 -0
- package/Dev/scripts/install-rust-rules.sh +139 -0
- package/Dev/scripts/install-tauri-rules.sh +149 -0
- package/Dev/stack-profiles/rust/permissions.txt +4 -0
- package/Dev/stack-profiles/rust/quality.md +3 -0
- package/Dev/stack-profiles/rust/stack.txt +2 -0
- package/Dev/stack-profiles/tauri/permissions.txt +7 -0
- package/Dev/stack-profiles/tauri/quality.md +4 -0
- package/Dev/stack-profiles/tauri/stack.txt +2 -0
- package/README.md +11 -5
- package/bundles/cursor/.cursorrules +5 -3
- package/bundles/windsurf/.windsurfrules +5 -3
- package/cli/lib/doctor.js +1 -1
- package/cli/lib/tech-registry.js +20 -0
- package/package.json +4 -7
|
@@ -0,0 +1,108 @@
|
|
|
1
|
+
# New Feature Development Checklist - Rust / Axum
|
|
2
|
+
|
|
3
|
+
## Before Starting
|
|
4
|
+
|
|
5
|
+
- [ ] Feature discussed and validated with the team
|
|
6
|
+
- [ ] Acceptance criteria written (Given/When/Then)
|
|
7
|
+
- [ ] Assumptions stated explicitly (inputs, invariants, volume, authorization, latency budget)
|
|
8
|
+
- [ ] API contract defined if exposed (OpenAPI 3.x, error format RFC 9457)
|
|
9
|
+
- [ ] Database changes identified (tables, columns, indexes, constraints)
|
|
10
|
+
- [ ] Impact analysis done: crates touched, breaking changes, rollback plan
|
|
11
|
+
- [ ] Feature flag decided for risky changes
|
|
12
|
+
|
|
13
|
+
## Development (TDD: Red, Green, Refactor)
|
|
14
|
+
|
|
15
|
+
### Design (layers first)
|
|
16
|
+
|
|
17
|
+
- [ ] Start from the domain: types, invariants and errors before routes
|
|
18
|
+
- [ ] Ports (traits) added only for real boundaries: repository, clock, mailer, external API
|
|
19
|
+
- [ ] One use case = one function or struct with one public method
|
|
20
|
+
- [ ] No speculative traits, generics, options, or parameters (YAGNI)
|
|
21
|
+
- [ ] Domain crate still free of axum, sqlx and tokio dependencies
|
|
22
|
+
|
|
23
|
+
### Domain Layer
|
|
24
|
+
|
|
25
|
+
- [ ] Newtypes validate on construction (`Email::parse`, `Quantity::new`); no primitive obsession
|
|
26
|
+
- [ ] States modeled as enums; illegal states unrepresentable where cheap
|
|
27
|
+
- [ ] Domain errors are a `thiserror` enum with one variant per failure the caller can act on
|
|
28
|
+
- [ ] No panics: fallible operations return `Result`
|
|
29
|
+
|
|
30
|
+
### Database
|
|
31
|
+
|
|
32
|
+
- [ ] Migration created with `sqlx migrate add` and is safe on large tables
|
|
33
|
+
- [ ] Foreign keys, indexes, `NOT NULL`, `UNIQUE` and `CHECK` constraints mirror domain invariants
|
|
34
|
+
- [ ] Queries written with `query!` / `query_as!`; dynamic filters through `QueryBuilder::push_bind`
|
|
35
|
+
- [ ] `cargo sqlx prepare --workspace` run and `.sqlx/` committed
|
|
36
|
+
- [ ] Multi-statement writes inside a transaction; no transaction held across external calls
|
|
37
|
+
|
|
38
|
+
### HTTP Layer
|
|
39
|
+
|
|
40
|
+
- [ ] Routes registered in the resource's `routes()` function; path syntax `/{id}` (axum 0.8)
|
|
41
|
+
- [ ] Handlers <= 15 lines: extract, validate, call the use case, map the result
|
|
42
|
+
- [ ] Request/response DTOs separate from domain types; `TryFrom<Dto>` builds validated domain values
|
|
43
|
+
- [ ] Authentication layer applied to the router; authorization checks the resource owner or tenant
|
|
44
|
+
- [ ] Rate limit added on sensitive endpoints (login, password reset, expensive operations)
|
|
45
|
+
- [ ] Errors flow through `AppError` and render RFC 9457 `application/problem+json`
|
|
46
|
+
- [ ] Correct statuses: 201 + `Location` on create, 204 on delete, 400/422 on validation, 404, 409, 403
|
|
47
|
+
- [ ] Body, timeout and CORS settings checked for the new route
|
|
48
|
+
|
|
49
|
+
### Async and Background Work
|
|
50
|
+
|
|
51
|
+
- [ ] No blocking calls on the runtime; CPU-bound work through `spawn_blocking`
|
|
52
|
+
- [ ] No lock guard held across `.await`
|
|
53
|
+
- [ ] Work slower than ~200 ms moved to a background worker or queue; jobs idempotent with retry policy
|
|
54
|
+
- [ ] Spawned tasks tracked and cancellation-safe; graceful shutdown still drains them
|
|
55
|
+
|
|
56
|
+
### Code Quality
|
|
57
|
+
|
|
58
|
+
- [ ] `cargo fmt` and `cargo clippy -D warnings` pass
|
|
59
|
+
- [ ] Borrowing preferred over cloning; iterators over index loops
|
|
60
|
+
- [ ] Naming explicit (RFC 430); comments explain WHY
|
|
61
|
+
- [ ] `tracing` spans/events added with structured fields; no secrets in logs
|
|
62
|
+
|
|
63
|
+
## Testing
|
|
64
|
+
|
|
65
|
+
- [ ] Domain tests: valid and invalid construction, state transitions, boundaries
|
|
66
|
+
- [ ] Use case tests against in-memory fakes: success and each failure variant
|
|
67
|
+
- [ ] HTTP tests with `tower::ServiceExt::oneshot` for every new endpoint, including 401 and 403 cases
|
|
68
|
+
- [ ] Database tests with `#[sqlx::test]`; migrations applied per test
|
|
69
|
+
- [ ] Error mapping tested (status, `problem+json` body)
|
|
70
|
+
- [ ] Property tests (`proptest`) for parsers and validators when relevant
|
|
71
|
+
- [ ] Coverage stays >= 80%; time injected; external HTTP stubbed
|
|
72
|
+
|
|
73
|
+
## Security
|
|
74
|
+
|
|
75
|
+
- [ ] `cargo audit` and `cargo deny check` clean
|
|
76
|
+
- [ ] No string-built SQL; no `unsafe` without `// SAFETY:` and tests
|
|
77
|
+
- [ ] IDOR checked: users cannot reach other users' or tenants' records
|
|
78
|
+
- [ ] Secrets in environment or secret store, wrapped in `SecretString`
|
|
79
|
+
- [ ] Input bounded (body size, collection sizes, pagination limits) and validated
|
|
80
|
+
- [ ] New logs do not contain sensitive data
|
|
81
|
+
|
|
82
|
+
## Performance
|
|
83
|
+
|
|
84
|
+
- [ ] No N+1 queries; batch with `ANY($1)` or joins
|
|
85
|
+
- [ ] Pagination on list endpoints (keyset where the table is large)
|
|
86
|
+
- [ ] Indexes support new filters and sorts; `EXPLAIN` checked on heavy queries
|
|
87
|
+
- [ ] Pool size and timeouts reviewed for the new load
|
|
88
|
+
|
|
89
|
+
## Documentation
|
|
90
|
+
|
|
91
|
+
- [ ] README or docs updated
|
|
92
|
+
- [ ] OpenAPI document updated
|
|
93
|
+
- [ ] ADR written for significant decisions
|
|
94
|
+
- [ ] CHANGELOG entry added; public items documented with `///`
|
|
95
|
+
|
|
96
|
+
## Before Merge
|
|
97
|
+
|
|
98
|
+
- [ ] `./scripts/ci.sh` passes locally and in CI
|
|
99
|
+
- [ ] Self-review of the diff completed
|
|
100
|
+
- [ ] PR description explains the why, links the ticket, lists risks and rollback
|
|
101
|
+
- [ ] Reviewer approval received; squash merge with Conventional Commit title
|
|
102
|
+
|
|
103
|
+
## Deployment
|
|
104
|
+
|
|
105
|
+
- [ ] Migrations run before the new binary starts (release step or startup migrate)
|
|
106
|
+
- [ ] Feature flag default set; rollback path verified
|
|
107
|
+
- [ ] Monitoring and alerts in place (error rate, latency, pool saturation)
|
|
108
|
+
- [ ] Smoke test after deploy (health endpoint and the main flow)
|
|
@@ -0,0 +1,118 @@
|
|
|
1
|
+
# Pre-Commit Checklist - Rust / Axum
|
|
2
|
+
|
|
3
|
+
## Quick Validation Before Each Commit
|
|
4
|
+
|
|
5
|
+
All commands run in Docker (`docker compose exec app ...`). Keep one script or `make ci` target that chains the steps below so the local gate equals the CI gate.
|
|
6
|
+
|
|
7
|
+
### Code Quality
|
|
8
|
+
|
|
9
|
+
- [ ] Formatting clean (`cargo fmt --all -- --check`)
|
|
10
|
+
- [ ] Clippy clean with warnings denied (`cargo clippy --workspace --all-targets --all-features -- -D warnings`)
|
|
11
|
+
- [ ] New `#[allow(...)]` justified, scoped to one item (prefer `#[expect(..., reason = "...")]`)
|
|
12
|
+
- [ ] No unused `use`, dead code, or commented-out code
|
|
13
|
+
- [ ] No `dbg!`, `println!`, `todo!` left behind
|
|
14
|
+
- [ ] Functions < 10 lines (limit 20), nesting <= 3, parameters <= 4
|
|
15
|
+
|
|
16
|
+
### Tests
|
|
17
|
+
|
|
18
|
+
- [ ] All tests pass (`cargo nextest run --workspace --all-features`)
|
|
19
|
+
- [ ] Doc tests pass (`cargo test --workspace --doc`)
|
|
20
|
+
- [ ] New behavior has tests written first (Red, Green, Refactor)
|
|
21
|
+
- [ ] Bug fixes include a regression test that failed before the fix
|
|
22
|
+
- [ ] Coverage maintained at 80% or higher (`cargo llvm-cov nextest --workspace --fail-under-lines 80`)
|
|
23
|
+
- [ ] No `#[ignore]` added without a ticket
|
|
24
|
+
|
|
25
|
+
### Security
|
|
26
|
+
|
|
27
|
+
- [ ] `cargo audit` reports no vulnerabilities
|
|
28
|
+
- [ ] `cargo deny check` passes (advisories, licenses, bans, sources)
|
|
29
|
+
- [ ] No hardcoded secrets; `.env` and key files not staged; secrets wrapped in `SecretString`
|
|
30
|
+
- [ ] `#![forbid(unsafe_code)]` still present; any new `unsafe` has a `// SAFETY:` comment and a test
|
|
31
|
+
- [ ] No `unwrap()` / `expect()` / indexing that can panic in request paths
|
|
32
|
+
- [ ] No string-built SQL (`format!` into a query); user input only bound as parameters
|
|
33
|
+
- [ ] New handlers authenticated, authorized per resource, and rate limited when they authenticate or send mail
|
|
34
|
+
|
|
35
|
+
### Database
|
|
36
|
+
|
|
37
|
+
- [ ] Migration added with `sqlx migrate add`, has indexes and foreign keys
|
|
38
|
+
- [ ] No data backfill mixed into the schema migration
|
|
39
|
+
- [ ] `cargo sqlx prepare --workspace` run and `.sqlx/` committed (`cargo sqlx prepare --check` passes)
|
|
40
|
+
- [ ] Migrations apply cleanly on an empty database
|
|
41
|
+
- [ ] Multi-statement writes wrapped in a transaction; no N+1 query loops in handlers
|
|
42
|
+
|
|
43
|
+
### Dependencies
|
|
44
|
+
|
|
45
|
+
- [ ] `Cargo.lock` committed and consistent (`cargo build --locked`)
|
|
46
|
+
- [ ] New crates justified (maintained, license allowed by `deny.toml`, no overlap with existing dependencies)
|
|
47
|
+
- [ ] Features trimmed (`default-features = false` where practical)
|
|
48
|
+
|
|
49
|
+
### Documentation and Git
|
|
50
|
+
|
|
51
|
+
- [ ] README / docs updated when behavior or setup changed
|
|
52
|
+
- [ ] CHANGELOG updated if applicable
|
|
53
|
+
- [ ] Commit message follows Conventional Commits (`feat(orders): add refund endpoint`)
|
|
54
|
+
- [ ] Commit is atomic; branch up to date with `main`
|
|
55
|
+
- [ ] No unrelated files staged (`git status`, `git diff --staged`)
|
|
56
|
+
|
|
57
|
+
## Automated Validation
|
|
58
|
+
|
|
59
|
+
### CI gate script
|
|
60
|
+
|
|
61
|
+
```bash
|
|
62
|
+
#!/usr/bin/env bash
|
|
63
|
+
# scripts/ci.sh - same steps locally and in CI
|
|
64
|
+
set -euo pipefail
|
|
65
|
+
|
|
66
|
+
cargo fmt --all -- --check
|
|
67
|
+
cargo clippy --workspace --all-targets --all-features -- -D warnings
|
|
68
|
+
cargo sqlx prepare --check --workspace
|
|
69
|
+
cargo nextest run --workspace --all-features
|
|
70
|
+
cargo test --workspace --doc
|
|
71
|
+
cargo deny check
|
|
72
|
+
cargo audit
|
|
73
|
+
```
|
|
74
|
+
|
|
75
|
+
Run it in the container:
|
|
76
|
+
|
|
77
|
+
```bash
|
|
78
|
+
docker compose exec app ./scripts/ci.sh
|
|
79
|
+
```
|
|
80
|
+
|
|
81
|
+
### Git pre-commit hook (Lefthook)
|
|
82
|
+
|
|
83
|
+
```yaml
|
|
84
|
+
# lefthook.yml
|
|
85
|
+
pre-commit:
|
|
86
|
+
parallel: true
|
|
87
|
+
commands:
|
|
88
|
+
fmt:
|
|
89
|
+
glob: "*.rs"
|
|
90
|
+
run: docker compose exec -T app cargo fmt --all -- --check
|
|
91
|
+
clippy:
|
|
92
|
+
glob: "*.{rs,toml}"
|
|
93
|
+
run: docker compose exec -T app cargo clippy --workspace --all-targets -- -D warnings
|
|
94
|
+
pre-push:
|
|
95
|
+
commands:
|
|
96
|
+
tests:
|
|
97
|
+
run: docker compose exec -T app cargo nextest run --workspace
|
|
98
|
+
deny:
|
|
99
|
+
run: docker compose exec -T app cargo deny check
|
|
100
|
+
```
|
|
101
|
+
|
|
102
|
+
Hooks are requirements, CLAUDE.md is a suggestion.
|
|
103
|
+
|
|
104
|
+
## Quick Commands
|
|
105
|
+
|
|
106
|
+
```bash
|
|
107
|
+
docker compose exec app cargo fmt --all
|
|
108
|
+
docker compose exec app cargo clippy --workspace --all-targets -- -D warnings
|
|
109
|
+
docker compose exec app cargo nextest run --workspace
|
|
110
|
+
docker compose exec app cargo llvm-cov nextest --workspace --summary-only
|
|
111
|
+
docker compose exec app cargo sqlx prepare --workspace
|
|
112
|
+
docker compose exec app cargo deny check
|
|
113
|
+
docker compose exec app cargo audit
|
|
114
|
+
```
|
|
115
|
+
|
|
116
|
+
## Final Check
|
|
117
|
+
|
|
118
|
+
If any box is unchecked, do not commit. Fix first, then re-run `./scripts/ci.sh`.
|
|
@@ -0,0 +1,251 @@
|
|
|
1
|
+
# Check Complete Rust / Axum Compliance
|
|
2
|
+
|
|
3
|
+
## Arguments
|
|
4
|
+
|
|
5
|
+
$ARGUMENTS (optional: path to project to analyze)
|
|
6
|
+
|
|
7
|
+
## Plan Mode
|
|
8
|
+
|
|
9
|
+
> Plan mode is activated automatically when the scope spans multiple crates or requires cross-cutting investigation.
|
|
10
|
+
|
|
11
|
+
## MISSION
|
|
12
|
+
|
|
13
|
+
Perform a complete compliance audit of the Rust / Axum project by orchestrating the 4 major checks: Architecture, Code Quality, Tests, and Security. Produce a consolidated report with an overall score out of 100 points.
|
|
14
|
+
|
|
15
|
+
### Step 1: Audit Preparation
|
|
16
|
+
|
|
17
|
+
Prepare audit environment:
|
|
18
|
+
- [ ] Identify project path to audit
|
|
19
|
+
- [ ] Verify presence of configuration files (`Cargo.toml`, `Cargo.lock`, `rust-toolchain.toml`, `deny.toml`, `rustfmt.toml`, `.sqlx/`)
|
|
20
|
+
- [ ] List main directories (`crates/` or `src/`, `migrations/`, `tests/`)
|
|
21
|
+
- [ ] Identify Rust toolchain, edition, and axum/tokio/sqlx versions
|
|
22
|
+
|
|
23
|
+
```bash
|
|
24
|
+
docker compose exec app rustc --version
|
|
25
|
+
docker compose exec app cargo --version
|
|
26
|
+
docker compose exec app cargo tree --workspace --depth 1 -i axum
|
|
27
|
+
docker compose exec app grep -nE "^(edition|rust-version)" Cargo.toml
|
|
28
|
+
```
|
|
29
|
+
|
|
30
|
+
**Note**: If $ARGUMENTS provided, use it as project path, otherwise use current directory.
|
|
31
|
+
|
|
32
|
+
### Step 2: Architecture Audit (25 points)
|
|
33
|
+
|
|
34
|
+
Execute architecture check: `/rust:check architecture` or follow `.claude/references/rust/checks/architecture.md`
|
|
35
|
+
|
|
36
|
+
**Evaluated Criteria**:
|
|
37
|
+
- Workspace layout and project structure (5 pts)
|
|
38
|
+
- Thin HTTP layer: routers, extractors, state (6 pts)
|
|
39
|
+
- Domain and application design, dependency-free domain (6 pts)
|
|
40
|
+
- Data layer: SQLx checked queries, migrations, transactions (4 pts)
|
|
41
|
+
- Async runtime hygiene, graceful shutdown, observability (2 pts)
|
|
42
|
+
- Dependency direction and configuration (2 pts)
|
|
43
|
+
|
|
44
|
+
**Reference**: `.claude/references/rust/checks/architecture.md`
|
|
45
|
+
|
|
46
|
+
### Step 3: Code Quality Audit (25 points)
|
|
47
|
+
|
|
48
|
+
Execute code quality check: `/rust:check code-quality` or follow `.claude/references/rust/checks/code-quality.md`
|
|
49
|
+
|
|
50
|
+
**Evaluated Criteria**:
|
|
51
|
+
- Clippy (`-D warnings`) and rustfmt clean (6 pts)
|
|
52
|
+
- Idiomatic Rust: borrowing, iterators, newtypes, no needless `unsafe` (5 pts)
|
|
53
|
+
- KISS/DRY/YAGNI and complexity (4 pts)
|
|
54
|
+
- Error handling: `thiserror`, `anyhow` only in binaries, no panics in request paths (4 pts)
|
|
55
|
+
- Naming conventions and documentation (3 pts)
|
|
56
|
+
- Toolchain pinning and CI (3 pts)
|
|
57
|
+
|
|
58
|
+
**Reference**: `.claude/references/rust/checks/code-quality.md`
|
|
59
|
+
|
|
60
|
+
### Step 4: Testing Audit (25 points)
|
|
61
|
+
|
|
62
|
+
Execute testing check: `/rust:check testing` or follow `.claude/references/rust/checks/testing.md`
|
|
63
|
+
|
|
64
|
+
**Evaluated Criteria**:
|
|
65
|
+
- Code coverage with cargo-llvm-cov, 80% minimum (7 pts)
|
|
66
|
+
- Domain and unit tests (6 pts)
|
|
67
|
+
- HTTP and integration tests (`oneshot`, `#[sqlx::test]`) (5 pts)
|
|
68
|
+
- Contract, end-to-end and doc tests (3 pts)
|
|
69
|
+
- Fixtures and test data (2 pts)
|
|
70
|
+
- Test isolation and speed (cargo-nextest) (2 pts)
|
|
71
|
+
|
|
72
|
+
**Reference**: `.claude/references/rust/checks/testing.md`
|
|
73
|
+
|
|
74
|
+
### Step 5: Security Audit (25 points)
|
|
75
|
+
|
|
76
|
+
Execute security check: `/rust:check security` or follow `.claude/references/rust/checks/security.md`
|
|
77
|
+
|
|
78
|
+
**Evaluated Criteria**:
|
|
79
|
+
- Memory safety and static analysis (`forbid(unsafe_code)`, clippy) (6 pts)
|
|
80
|
+
- Dependency vulnerabilities and supply chain (cargo-audit, cargo-deny) (4 pts)
|
|
81
|
+
- Secrets and credentials management (4 pts)
|
|
82
|
+
- Authentication and authorization (Argon2id, JWT pinning, IDOR) (4 pts)
|
|
83
|
+
- Injection and input validation (4 pts)
|
|
84
|
+
- Secure configuration, headers, rate limiting (3 pts)
|
|
85
|
+
|
|
86
|
+
**Reference**: `.claude/references/rust/checks/security.md`
|
|
87
|
+
|
|
88
|
+
### Step 6: Consolidation and Global Scoring
|
|
89
|
+
|
|
90
|
+
- [ ] Sum the 4 scores (max 100 points)
|
|
91
|
+
- [ ] Identify critical categories (<50%)
|
|
92
|
+
- [ ] List all critical cross-cutting issues
|
|
93
|
+
- [ ] Prioritize actions by impact/effort
|
|
94
|
+
- [ ] Produce final consolidated report
|
|
95
|
+
|
|
96
|
+
**Grading Scale**:
|
|
97
|
+
- 90-100: Excellent - Reference project
|
|
98
|
+
- 75-89: Very Good - Some minor improvements
|
|
99
|
+
- 60-74: Acceptable - Requires improvements
|
|
100
|
+
- 40-59: Insufficient - Major refactoring required
|
|
101
|
+
- 0-39: Critical - Complete overhaul necessary
|
|
102
|
+
|
|
103
|
+
### Step 7: Recommendations and Action Plan
|
|
104
|
+
|
|
105
|
+
- [ ] Identify top 3 priority actions across all categories
|
|
106
|
+
- [ ] Estimate effort (Low/Medium/High) and impact (Low/Medium/High) for each
|
|
107
|
+
- [ ] Propose implementation order
|
|
108
|
+
- [ ] Suggest quick wins (high impact/effort ratio)
|
|
109
|
+
|
|
110
|
+
**Version health** (reported in the executive summary, not scored):
|
|
111
|
+
- Rust stable is the baseline; flag a toolchain more than a few releases behind stable and an MSRV that blocks security updates.
|
|
112
|
+
- Edition 2024 is expected for new code; older editions are accepted with a migration plan (`cargo fix --edition`).
|
|
113
|
+
- axum 0.8, tokio 1.x and sqlx 0.9 are the reference lines; flag older minor lines and any yanked crate in `Cargo.lock`.
|
|
114
|
+
|
|
115
|
+
## OUTPUT FORMAT
|
|
116
|
+
|
|
117
|
+
```
|
|
118
|
+
RUST / AXUM COMPLIANCE AUDIT - COMPLETE REPORT
|
|
119
|
+
==============================================
|
|
120
|
+
|
|
121
|
+
OVERALL SCORE: XX/100
|
|
122
|
+
|
|
123
|
+
COMPLIANCE LEVEL: [Excellent/Very Good/Acceptable/Insufficient/Critical]
|
|
124
|
+
|
|
125
|
+
Rust: X.Y.Z Edition: 20XX axum: X.Y Test runner: cargo-nextest
|
|
126
|
+
|
|
127
|
+
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
|
|
128
|
+
|
|
129
|
+
SCORES BY CATEGORY:
|
|
130
|
+
|
|
131
|
+
ARCHITECTURE : XX/25 [██████████░░░░░░░░░░] XX%
|
|
132
|
+
CODE QUALITY : XX/25 [██████████░░░░░░░░░░] XX%
|
|
133
|
+
TESTS : XX/25 [██████████░░░░░░░░░░] XX%
|
|
134
|
+
SECURITY : XX/25 [██████████░░░░░░░░░░] XX%
|
|
135
|
+
|
|
136
|
+
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
|
|
137
|
+
|
|
138
|
+
OVERALL STRENGTHS:
|
|
139
|
+
1. [Strength identified in multiple categories]
|
|
140
|
+
2. [Other major strength]
|
|
141
|
+
3. [Third strength]
|
|
142
|
+
|
|
143
|
+
OVERALL IMPROVEMENTS:
|
|
144
|
+
1. [Cross-cutting improvement]
|
|
145
|
+
2. [Other recommended improvement]
|
|
146
|
+
3. [Third improvement]
|
|
147
|
+
|
|
148
|
+
CRITICAL ISSUES:
|
|
149
|
+
1. [Critical issue #1 - affected category]
|
|
150
|
+
2. [Critical issue #2 - affected category]
|
|
151
|
+
3. [Critical issue #3 - affected category]
|
|
152
|
+
|
|
153
|
+
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
|
|
154
|
+
|
|
155
|
+
DETAILS BY CATEGORY:
|
|
156
|
+
|
|
157
|
+
┌─────────────────────────────────────────────┐
|
|
158
|
+
│ ARCHITECTURE (XX/25) │
|
|
159
|
+
└─────────────────────────────────────────────┘
|
|
160
|
+
|
|
161
|
+
Sub-scores:
|
|
162
|
+
• Workspace and structure : XX/5
|
|
163
|
+
• Thin HTTP layer : XX/6
|
|
164
|
+
• Domain / application design : XX/6
|
|
165
|
+
• Data layer : XX/4
|
|
166
|
+
• Async runtime / operations : XX/2
|
|
167
|
+
• Dependency direction : XX/2
|
|
168
|
+
|
|
169
|
+
Strengths:
|
|
170
|
+
- [Architecture strengths]
|
|
171
|
+
|
|
172
|
+
Issues:
|
|
173
|
+
- [Architecture issues]
|
|
174
|
+
|
|
175
|
+
[Similar sections for Code Quality, Tests, and Security...]
|
|
176
|
+
|
|
177
|
+
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
|
|
178
|
+
|
|
179
|
+
TOP 3 PRIORITY ACTIONS (ALL CATEGORIES):
|
|
180
|
+
|
|
181
|
+
1. CRITICAL - [Action #1]
|
|
182
|
+
Category : [Architecture/Quality/Tests/Security]
|
|
183
|
+
Impact : [High/Medium/Low]
|
|
184
|
+
Effort : [High/Medium/Low]
|
|
185
|
+
Priority : IMMEDIATE
|
|
186
|
+
|
|
187
|
+
Detailed description:
|
|
188
|
+
[Explanation of problem and proposed solution]
|
|
189
|
+
|
|
190
|
+
Affected files:
|
|
191
|
+
- [file:line]
|
|
192
|
+
|
|
193
|
+
Correction example:
|
|
194
|
+
[Code or correction command]
|
|
195
|
+
|
|
196
|
+
2. IMPORTANT - [Action #2]
|
|
197
|
+
[Same format...]
|
|
198
|
+
|
|
199
|
+
3. RECOMMENDED - [Action #3]
|
|
200
|
+
[Same format...]
|
|
201
|
+
|
|
202
|
+
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
|
|
203
|
+
|
|
204
|
+
QUICK WINS (High Impact / Low Effort):
|
|
205
|
+
|
|
206
|
+
- [Quick win #1] - Category: [X] - Impact: [X] - Effort: [X]
|
|
207
|
+
- [Quick win #2] - Category: [X] - Impact: [X] - Effort: [X]
|
|
208
|
+
- [Quick win #3] - Category: [X] - Impact: [X] - Effort: [X]
|
|
209
|
+
|
|
210
|
+
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
|
|
211
|
+
|
|
212
|
+
RECOMMENDED ACTION PLAN:
|
|
213
|
+
|
|
214
|
+
WEEK 1 (Immediate):
|
|
215
|
+
- [ ] [Critical action #1]
|
|
216
|
+
- [ ] [Priority quick win]
|
|
217
|
+
|
|
218
|
+
WEEK 2-4 (Short term):
|
|
219
|
+
- [ ] [Important action #2]
|
|
220
|
+
- [ ] [Other quick wins]
|
|
221
|
+
|
|
222
|
+
MONTH 2-3 (Medium term):
|
|
223
|
+
- [ ] [Recommended action #3]
|
|
224
|
+
- [ ] [Progressive improvements]
|
|
225
|
+
|
|
226
|
+
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
|
|
227
|
+
|
|
228
|
+
EXECUTIVE SUMMARY:
|
|
229
|
+
|
|
230
|
+
[Summary paragraph on overall project state, major strengths,
|
|
231
|
+
major weaknesses, version health (toolchain, edition and key crate
|
|
232
|
+
lines), and recommended trajectory. Say whether the project is
|
|
233
|
+
production-ready, requires corrections, or needs refactoring.]
|
|
234
|
+
|
|
235
|
+
General Recommendation: [Production-ready / Minor corrections /
|
|
236
|
+
Major refactoring / Overhaul necessary]
|
|
237
|
+
|
|
238
|
+
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
|
|
239
|
+
```
|
|
240
|
+
|
|
241
|
+
## IMPORTANT NOTES
|
|
242
|
+
|
|
243
|
+
- This command orchestrates the 4 specialized audits
|
|
244
|
+
- Use Docker for all analysis tools (`docker compose exec app ...`)
|
|
245
|
+
- Provide concrete examples with file:line for each problem
|
|
246
|
+
- Prioritize actions based on Impact/Effort matrix
|
|
247
|
+
- Security problems are ALWAYS top priority
|
|
248
|
+
- A single-crate service with clean modules is acceptable; do not penalize the absence of a workspace on its own
|
|
249
|
+
- Propose automatable corrections (CI workflow, pre-commit hooks, `cargo deny`, `cargo clippy --fix` after review)
|
|
250
|
+
- Report must be actionable, not just descriptive
|
|
251
|
+
- Adapt recommendations to project business context
|
|
@@ -0,0 +1,182 @@
|
|
|
1
|
+
# Check Rust / Axum Architecture
|
|
2
|
+
|
|
3
|
+
## Arguments
|
|
4
|
+
|
|
5
|
+
$ARGUMENTS (optional: path to project to analyze)
|
|
6
|
+
|
|
7
|
+
## Plan Mode
|
|
8
|
+
|
|
9
|
+
> Plan mode is activated automatically when the scope spans multiple crates or requires cross-cutting investigation.
|
|
10
|
+
|
|
11
|
+
## MISSION
|
|
12
|
+
|
|
13
|
+
Audit the architecture of the Rust web service (Rust stable, edition 2024, axum 0.8, tokio 1.x, sqlx 0.9). Stance: **layered or hexagonal, enforced by the crate graph**. The domain crate must not depend on axum, sqlx or tokio; the HTTP layer adapts requests to use cases and nothing more. A single crate with well-separated modules is acceptable for small services; do not demand a workspace the project does not need (YAGNI).
|
|
14
|
+
|
|
15
|
+
All commands run in Docker (`docker compose exec app ...`).
|
|
16
|
+
|
|
17
|
+
### Step 1: Workspace and Project Structure (5 points)
|
|
18
|
+
|
|
19
|
+
```bash
|
|
20
|
+
docker compose exec app cargo metadata --no-deps --format-version 1 | jq -r '.packages[].name'
|
|
21
|
+
docker compose exec app cargo tree --workspace --depth 1
|
|
22
|
+
docker compose exec app ls crates/ migrations/ .sqlx/
|
|
23
|
+
docker compose exec app grep -nE "edition|rust-version|resolver|\[workspace" Cargo.toml
|
|
24
|
+
```
|
|
25
|
+
|
|
26
|
+
Checklist:
|
|
27
|
+
- [ ] Cargo workspace with one crate per layer (for example `domain`, `application`, `infrastructure`, `http`, plus a thin `app` binary), or a single crate with equivalent top-level modules
|
|
28
|
+
- [ ] `edition = "2024"` and `rust-version` (MSRV) declared; `resolver = "3"` or the edition default
|
|
29
|
+
- [ ] Shared dependency versions centralized in `[workspace.dependencies]`; shared lints in `[workspace.lints]`
|
|
30
|
+
- [ ] `Cargo.lock` committed for binaries; `rust-toolchain.toml` pins the channel
|
|
31
|
+
- [ ] The binary crate only wires dependencies (config, pool, router, server); no business code in `main.rs`
|
|
32
|
+
- [ ] SQL migrations in `migrations/` (sqlx-cli) and the offline query cache `.sqlx/` committed
|
|
33
|
+
|
|
34
|
+
Detect anti-patterns:
|
|
35
|
+
|
|
36
|
+
```bash
|
|
37
|
+
# Giant files
|
|
38
|
+
docker compose exec app sh -c 'find crates src -name "*.rs" | xargs wc -l | sort -rn | head -10'
|
|
39
|
+
# Business code in main
|
|
40
|
+
docker compose exec app sh -c 'wc -l src/main.rs crates/*/src/main.rs'
|
|
41
|
+
```
|
|
42
|
+
|
|
43
|
+
**Scoring**: 5 pts, minus 1 per violation (min 0).
|
|
44
|
+
|
|
45
|
+
### Step 2: Thin HTTP Layer (6 points)
|
|
46
|
+
|
|
47
|
+
Handlers coordinate: extract, validate, call a use case, map the result to a response. No business rules, no SQL.
|
|
48
|
+
|
|
49
|
+
```bash
|
|
50
|
+
docker compose exec app sh -c 'find crates src -path "*http*" -name "*.rs" | xargs wc -l | sort -rn | head -10'
|
|
51
|
+
# SQL or pool access inside handlers
|
|
52
|
+
docker compose exec app grep -rnE "sqlx::query|\.fetch_(one|all|optional)|\.execute\(" crates/http src/http src/handlers
|
|
53
|
+
# Panics in request paths
|
|
54
|
+
docker compose exec app grep -rnE "\.unwrap\(\)|\.expect\(|panic!|unreachable!|todo!|unimplemented!" crates/http src/http src/handlers
|
|
55
|
+
```
|
|
56
|
+
|
|
57
|
+
Checklist:
|
|
58
|
+
- [ ] Handlers are <= 15 lines (target 10): extractors in, use case call, response out
|
|
59
|
+
- [ ] Router assembled in one function per resource (`fn routes() -> Router<AppState>`) and merged/nested in the app builder
|
|
60
|
+
- [ ] Path parameters use the 0.8 syntax (`/{id}`, `/{*rest}`), not the removed `/:id`
|
|
61
|
+
- [ ] Shared state is a cheap-to-clone `AppState` (`Arc` inside, or `Clone` handles such as `PgPool`) injected with `State<AppState>`; no global mutable statics
|
|
62
|
+
- [ ] Request/response DTOs are separate from domain types and database rows
|
|
63
|
+
- [ ] Validation at the boundary (`validator`, or `TryFrom<Dto>` building validated domain types); no `unwrap()`/`expect()` in request paths
|
|
64
|
+
- [ ] Cross-cutting concerns (tracing, timeout, CORS, body limit, auth) are Tower layers, not copy-pasted in handlers
|
|
65
|
+
- [ ] Custom extractors implement `FromRequestParts` / `FromRequest` without `#[async_trait]` (not needed in axum 0.8)
|
|
66
|
+
|
|
67
|
+
**Scoring**: 6 pts. Deduct 1 per handler with SQL or business rules, 1 per `unwrap`/`expect` cluster in request paths, max -6.
|
|
68
|
+
|
|
69
|
+
### Step 3: Domain and Application Design (6 points)
|
|
70
|
+
|
|
71
|
+
```bash
|
|
72
|
+
docker compose exec app cat crates/domain/Cargo.toml
|
|
73
|
+
docker compose exec app cargo tree -p domain --edges normal --depth 1
|
|
74
|
+
docker compose exec app grep -rnE "^use (axum|sqlx|tokio|tower|hyper)" crates/domain/src
|
|
75
|
+
docker compose exec app grep -rnE "(pub )?trait \w+" crates/application/src crates/domain/src
|
|
76
|
+
```
|
|
77
|
+
|
|
78
|
+
Checklist:
|
|
79
|
+
- [ ] The domain crate has no dependency on axum, sqlx, tokio or tower
|
|
80
|
+
- [ ] Invariants live in types: newtypes (`Email`, `UserId`) validated in constructors; enums for states instead of strings or booleans
|
|
81
|
+
- [ ] Ports (repository, clock, mailer traits) are defined in domain/application; adapters live in infrastructure
|
|
82
|
+
- [ ] Use cases are plain async functions or structs with a single public method; dependencies injected through constructors or generics
|
|
83
|
+
- [ ] Trait objects (`dyn`) used only where runtime polymorphism is needed; generics preferred otherwise
|
|
84
|
+
- [ ] Domain errors are explicit enums (`thiserror`), not `anyhow::Error`
|
|
85
|
+
- [ ] Illegal states are unrepresentable (typestate or enums) where cheap
|
|
86
|
+
- [ ] No God types: a struct named `Manager`, `Service` or `Utils` with many unrelated methods is a smell
|
|
87
|
+
|
|
88
|
+
**Scoring**: 6 pts. Deduct 2 if the domain depends on axum/sqlx, 1 per use case mixing HTTP or SQL concerns, 1 per god type.
|
|
89
|
+
|
|
90
|
+
### Step 4: Data Layer (4 points)
|
|
91
|
+
|
|
92
|
+
```bash
|
|
93
|
+
docker compose exec app grep -rnE "query!|query_as!|query_scalar!" crates src | wc -l
|
|
94
|
+
docker compose exec app grep -rnE "format!\(.*(SELECT|INSERT|UPDATE|DELETE)" crates src
|
|
95
|
+
docker compose exec app cargo sqlx prepare --check --workspace
|
|
96
|
+
docker compose exec app sqlx migrate info
|
|
97
|
+
docker compose exec app grep -rnE "PgPoolOptions|max_connections|acquire_timeout" crates src
|
|
98
|
+
```
|
|
99
|
+
|
|
100
|
+
Checklist:
|
|
101
|
+
- [ ] Queries use `query!` / `query_as!` (compile-time checked) with the offline cache; `cargo sqlx prepare --check` passes in CI
|
|
102
|
+
- [ ] SQL is never built by string concatenation or `format!`; dynamic filters use `sqlx::QueryBuilder` with `push_bind`
|
|
103
|
+
- [ ] Every foreign key has an index and a database constraint; `NOT NULL`, `UNIQUE` and `CHECK` mirror domain invariants
|
|
104
|
+
- [ ] Migrations are forward-only and safe on large tables (concurrent index, backfill in a separate step)
|
|
105
|
+
- [ ] One `PgPool` created at startup with explicit `max_connections` and `acquire_timeout`; migrations run at startup or as a deploy step
|
|
106
|
+
- [ ] Multi-statement writes use a transaction (`pool.begin()`), committed explicitly; no transaction held across external calls
|
|
107
|
+
- [ ] Repositories return domain types, not `sqlx::Row`; row structs stay in infrastructure
|
|
108
|
+
|
|
109
|
+
**Scoring**: 4 pts, minus 1 per failed group above (checked queries, schema integrity, pool and migrations, transactions).
|
|
110
|
+
|
|
111
|
+
### Step 5: Async Runtime and Operations (2 points)
|
|
112
|
+
|
|
113
|
+
```bash
|
|
114
|
+
docker compose exec app grep -rnE "std::thread::sleep|std::fs::|reqwest::blocking" crates src
|
|
115
|
+
docker compose exec app grep -rnE "spawn_blocking|tokio::spawn|JoinSet" crates src
|
|
116
|
+
docker compose exec app grep -rnE "with_graceful_shutdown|ctrl_c|SignalKind" crates src
|
|
117
|
+
docker compose exec app grep -rnE "TraceLayer|TimeoutLayer|tracing_subscriber" crates src
|
|
118
|
+
```
|
|
119
|
+
|
|
120
|
+
Checklist:
|
|
121
|
+
- [ ] No blocking calls on the async runtime; CPU-bound or blocking work goes through `tokio::task::spawn_blocking`
|
|
122
|
+
- [ ] No `std::sync::Mutex` guard held across an `.await`
|
|
123
|
+
- [ ] Graceful shutdown: `axum::serve(listener, app).with_graceful_shutdown(signal)` handling Ctrl+C and SIGTERM
|
|
124
|
+
- [ ] Spawned tasks are tracked (`JoinSet`, handles) and their errors observed; no fire-and-forget panics
|
|
125
|
+
- [ ] Health endpoints for liveness and readiness (database reachable)
|
|
126
|
+
- [ ] Structured logs through `tracing` with request ids (`TraceLayer`); subscriber initialized once in `main`
|
|
127
|
+
|
|
128
|
+
**Scoring**: 2 pts (1 for runtime hygiene, 1 for shutdown and observability).
|
|
129
|
+
|
|
130
|
+
### Step 6: Dependency Direction and Features (2 points)
|
|
131
|
+
|
|
132
|
+
```bash
|
|
133
|
+
docker compose exec app cargo tree --workspace --duplicates | head -30
|
|
134
|
+
docker compose exec app cargo tree -i axum --workspace
|
|
135
|
+
docker compose exec app grep -rnE "std::env::var|dotenv" crates src
|
|
136
|
+
```
|
|
137
|
+
|
|
138
|
+
Checklist:
|
|
139
|
+
- [ ] Dependencies point inward: http and infrastructure depend on application and domain, never the reverse; no cycles
|
|
140
|
+
- [ ] Feature flags of heavy crates are trimmed (`default-features = false` where practical)
|
|
141
|
+
- [ ] Configuration parsed once at startup into a typed struct; `std::env::var` is not scattered through the code
|
|
142
|
+
- [ ] Duplicated major versions of the same crate are reviewed
|
|
143
|
+
|
|
144
|
+
**Scoring**: 2 pts (1 for dependency direction, 1 for configuration and dependency hygiene).
|
|
145
|
+
|
|
146
|
+
## OUTPUT FORMAT
|
|
147
|
+
|
|
148
|
+
```
|
|
149
|
+
RUST / AXUM ARCHITECTURE AUDIT
|
|
150
|
+
==============================
|
|
151
|
+
|
|
152
|
+
SCORE: XX/25
|
|
153
|
+
|
|
154
|
+
Rust: X.Y.Z Edition: 20XX axum: X.Y Crates: N Handlers: N
|
|
155
|
+
|
|
156
|
+
SUB-SCORES:
|
|
157
|
+
Workspace and structure : XX/5
|
|
158
|
+
Thin HTTP layer : XX/6
|
|
159
|
+
Domain / application design : XX/6
|
|
160
|
+
Data layer : XX/4
|
|
161
|
+
Async runtime / operations : XX/2
|
|
162
|
+
Dependency direction : XX/2
|
|
163
|
+
|
|
164
|
+
STRENGTHS:
|
|
165
|
+
- [...]
|
|
166
|
+
|
|
167
|
+
ISSUES (by severity):
|
|
168
|
+
- CRITICAL: [file:line] [description] -> [fix]
|
|
169
|
+
- IMPORTANT: ...
|
|
170
|
+
- MINOR: ...
|
|
171
|
+
|
|
172
|
+
RECOMMENDATIONS:
|
|
173
|
+
1. [Action] (effort: Low/Medium/High, impact: Low/Medium/High)
|
|
174
|
+
2. ...
|
|
175
|
+
```
|
|
176
|
+
|
|
177
|
+
## IMPORTANT NOTES
|
|
178
|
+
|
|
179
|
+
- Findings must cite `file:line` and propose a concrete refactoring (for example "move the pricing rule from `create_order` into `Order::place` in the domain crate")
|
|
180
|
+
- Do not recommend new crates or traits when the code is small and readable (YAGNI); one implementation does not need a trait until a second one or a test double exists
|
|
181
|
+
- A single-crate service with clean modules scores fully on layering if the dependency direction holds
|
|
182
|
+
- Never run destructive tasks (`sqlx database drop`, `sqlx migrate revert`) during an audit
|