Skip to content
Skills Library
Programming · Review & audit

smrt-code-review

/smrt-code-review

Use when reviewing a downstream SMRT project. Fetch deterministic SMRT and HappyVertical SDK context from smrt-dev-mcp, inspect the actual code diff, and produce a findings-first code review.

Repository
smrt ↗
Licence
MIT
Last repository change
27.10.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
2
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 27.10.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Use when reviewing a downstream SMRT project. Fetch deterministic SMRT and HappyVertical SDK context from smrt-dev-mcp, inspect the actual code diff, and produce a findings-first code review.

Original author description

Use when reviewing a downstream SMRT project. Fetch deterministic SMRT and HappyVertical SDK context from smrt-dev-mcp, inspect the actual code diff, and produce a findings-first code review.

How it works

  1. Workflow
  2. Required Review Focus
  3. Output
Original SKILL.md headings · 65 lines, 371 words
SKILL.md preview
---
name: smrt-code-review
description: Use when reviewing a downstream SMRT project. Fetch deterministic SMRT and HappyVertical SDK context from smrt-dev-mcp, inspect the actual code diff, and produce a findings-first code review.
---

# SMRT Review

Use this skill for code reviews in projects that use SMRT packages or generated
SMRT surfaces.

## Workflow

1. Determine the repository root and changed files.
   - Prefer the user's explicit changed-file list.
   - Otherwise use the active harness's git support or `git diff --name-only`.
   - Include staged files when the user is preparing a commit.
2. Call MCP tool `build-context` from `smrt-dev-mcp` with `task: "review"` and:

```json
{
  "rootDir": "/absolute/path/to/project",
  "changedFiles": ["relative/path/from/root.ts"],
  "focus": "Review SMRT usage, relationships-v2 patterns, tenancy, generated API/CLI/MCP compatibility, SDK usage, prompt/data safety, and stale docs.",
  "mode": "both"
}
```

3. Treat `deterministicFindings` as required review routing, not as the whole
   review. A routing warning means the code needs focused scrutiny; do not
   report it as a defect unless the actual diff confirms a concrete issue.
4. Read the actual changed diff and any referenced files needed to understand
   behavior. Do not review from the MCP context alone.
5. Use `promptBundle.contextMarkdown` as current SMRT ecosystem context for
   selected packages and SDKs. When generated facts conflict with authored
   guidance, trust generated facts and flag the doc drift.
6. Produce a findings-first review. Prioritize correctness, regressions,
   missing tests, generated-surface compatibility, relationship/tenant
   invariants, data exposure, and stale docs.
7. If code, docs, or package expertise changed, call MCP tool
   `check-knowledge-freshness` with `{ "strict": true }` when available, or
   tell the user to run:

```bash
pnpm knowledge:check --strict --format markdown
```

## Required Review Focus

- Relationships-v2: `@foreignKey`, `@crossPackageRef`, `SmrtJunction`,
  `SmrtHierarchical`, polymorphic `(metaType, metaId, role)` links, qualified
  `_meta_type`, UUID id/FK columns, and tenant-guarded relationship loads.
- Generated surfaces: public exports, CLI commands, REST handlers, generated MCP
  tool names/schemas, and AI-operation compatibility.
- Tenancy: nullable tenant semantics, tenant context propagation, and explicit
  cross-tenant exceptions.
- SDK usage: prefer HappyVertical SDK packages for AI, SQL, files, logging,
  secrets, jobs, and external capability adapters.
- Agent docs: `AGENTS.md` is canonical; `CLAUDE.md` must remain exactly the
  `@AGENTS.md` shim.

## Output

Lead with findings ordered by severity. Use file and line references when the
harness can provide them. If no issues are found, say that clearly and mention
the validation or residual risk. Keep summaries secondary to findings.

See `references/review-output.md` for a stricter output template when the user
asks for a formal review.
Files in this package (2)
  • SKILL.md 3 KB
  • references/review-output.md 1 KB

Versions

The database has no other skills named smrt-code-review.

Database snapshot: 13.11.2026 · package version 20a02e8f4d92 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 90smrt-contribute
    Use when working in a fork or clone of the happyvertical/smrt repository as an outside contributor, for example "fix a bug in smrt-jobs and open a PR", "contribute this change upstream", or "add docs to smrt-core". Gives the fork pull-request procedure from CONTRIBUTING.md and…
  • 90kitchen
    Plan a business app with SMRT recipes in conversation, validate it, and create the project in a folder the user confirms. Use when the user wants to design, scope or start an app (a shop, studio, workshop, bakery...) on SMRT, or asks what SMRT can do for their business.
  • 90review
    Review a pull request or diff for this repository against otari's project standards. Use when asked to review a PR, inspect a diff, or prepare review feedback.
  • 90om-code-review
    Use when a diff, branch, or PR needs a correctness and security review, e.g. 'review this diff'. Run validation and the built-in plus repository checklist, check breaking changes and quality, and return severity-ranked findings with an approve/request-changes verdict. Powers…
  • 100code-review
    When the user asks for a code review, shares code for feedback, or says "review this", "check my code", "what's wrong with this". Also activate when reviewing a pull request or diff.
  • 85github-review
    Review a specified GitHub pull request at its current head, inspect the diff and surrounding context, and prepare actionable findings with check status. Use for PR review; reviewing does not authorize comments, approval, pushes, or merging.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.