Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/evgenii-studitskikh/claude-code-saas-studio/code-reviewnpx skills add evgenii-studitskikh/Claude-Code-SaaS-Studio --skill code-reviewgit clone --depth 1 https://github.com/evgenii-studitskikh/Claude-Code-SaaS-StudioWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/evgenii-studitskikh/claude-code-saas-studio/code-review)<a href="https://agentmods.dev/skills/evgenii-studitskikh/claude-code-saas-studio/code-review"><img src="https://agentmods.dev/badge/skills/evgenii-studitskikh/claude-code-saas-studio/code-review.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00054 | $0.00904 |
| Opus 5 | $0.00027 | $0.00452 |
| Sonnet 5 | $0.00011 | $0.00181 |
| Haiku 4.5 | $0.00005 | $0.00090 |
Grade A, and why
code-review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 25 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Perform a structured, rule-driven review of the current working-tree diff, checking for the most critical SaaS security and correctness properties. Non-autonomous: all findings are reported as a prioritized list; nothing is changed without the user's explicit sign-off on each fix.
Phases
- Get the diff — read the current unstaged and staged changes. If no diff exists, ask the user which files or commits to review. Summarize the scope of the review (N files, N lines changed) before proceeding.
- Check for secrets — scan all changed files for patterns that look like hardcoded secrets: API keys, JWTs, connection strings, private keys, and Stripe secret keys. Any match is a BLOCKING finding. Cross-check that
.env.exampleuses placeholders and that.gitignorecovers.env*. - Verify RLS and tenant scoping — for any server action or route handler, confirm that: (a) the Supabase client is initialized with the session cookie (not the service role key unless justified), (b) all queries that touch tenant data include a tenant-scoping filter, and (c) tenant ID is never sourced from unvalidated client input.
- Validate Stripe handling — for any webhook handler, verify: (a)
stripe.webhooks.constructEventis called with the raw request body andSTRIPE_WEBHOOK_SECRET, (b) event types are handled with idempotency (check for a processed-event record or Supabase upsert pattern), and (c) the handler returns 200 quickly and does heavy work asynchronously if needed. - Check input validation — confirm that all server actions and route handlers validate inputs with Zod (or equivalent) before any database or Stripe call. Note any handler that accepts
anytyped input or skips validation. - QA coverage advisory (qa-engineer perspective) — scan the list of changed source files and check whether each has a corresponding test file (
*.test.*,*.spec.*, or a matching path undere2e/). For every changed file with no accompanying test, note it as a NOTE-level finding with the suggestion: "Consider running/write-teststo add coverage for this file." This check is advisory only — it does not block the review verdict. - Security pass advisory (security-engineer perspective) — perform a quick, targeted scan of the diff for the following patterns, each reported as a WARNING unless it overlaps with an already-BLOCKING finding from earlier phases: (a) any token-shaped string that looks like a committed secret not caught by phase 2, (b) server actions or route handlers that read tenant data without an explicit RLS or tenant-scoping guard, (c) server actions or route handlers that accept inputs without Zod validation, and (d) Stripe webhook handlers missing
stripe.webhooks.constructEvent. Skip any issue already reported in phases 2–5; this advisory only surfaces NEW test-coverage or security observations not already covered. Close this phase with: "For a comprehensive security review run/security-audit." This advisory does not auto-fix or block; all findings are reported for human review. - Report findings — present findings grouped by severity: BLOCKING (must fix before ship), WARNING (should fix), and NOTE (consider fixing). For each finding, include the file path, line reference, and a one-sentence fix suggestion. Under
fullreview, discuss each finding; underlean/solo, present the list and ask for a single go/no-go to apply suggested fixes. Once all findings are resolved or acknowledged, return to the skill that invoked this review (e.g.,/build-feature) or continue the current phase.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 25 lines · 54 tokens per session scan A 0f60e0510551
code-review is a skill published in the GitHub repository evgenii-studitskikh/Claude-Code-SaaS-Studio (1 stars, last pushed 2mo ago), licensed MIT. It adds 54 tokens to every session and 904 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
create-modal
Create declarative modals using the modal library API. Covers modal types (confirm, input, select, form), sections (Text, Buttons, Input, Textarea, Checkbox, List, Combo, When, Custom), rendering with OverlayModal, and keyboard/mouse handling. Use when adding modals or dialogs to the application.
autoprompt
Explicit-only useful-first orchestration. Invoke /autoprompt to turn a mission into one executable roadmap, build dependency-safe lanes, and verify the result with independent reviewers. Never infer invocation from ordinary requests. Never resume from leftover artifacts without an explicit resume instruction.
openlore-brainstorm
Transform a feature idea into an annotated story using a Domain Sketch or Constrained Option Tree. Use when asked to brainstorm, explore, or shape a feature before implementation.
gemigo-cli
Use when the user wants to publish an already-built static website or front-end app through GemiGo, such as a Vite/React/Vue static build, plain HTML/CSS/JS page, landing page, demo, docs site, or small browser app, and get a hosted public URL.
ap-juror
L4 terminal leaf - G7 SIGN-OFF. One independent sign-off panel seat that saw none of the intermediate work. Binary PASS/FAIL on opened evidence; default-FAIL. A FAIL naming a P0/P1 blocker is NOT arbitrable into PASS.
agenticx-query-data-source
Use when the user asks about verifiable quantitative facts (stock prices, financial indicators, macro data, company registry, academic metrics, legal statutes) that must come from a live data source rather than training memory.