code-ai-installer 1.2.0 → 1.3.0

package/locales/en/.agents/security_review/SKILL.md

@@ -1,55 +1,277 @@
----
-name: security_review
-description: AppSec review for endpoints/authorization/input/secrets/sensitive features. Checklist + patterns (OWASP baseline).
----
-
-# Skill: Security Review (AppSec)
-
-## When to activate
-- New API endpoints
-- AuthN/AuthZ
-- User input / file uploads
-- Secrets/credentials
-- Payments/PII/sensitive data
-- Third-party integrations
-
-## Checklist (minimum)
-1) Secrets management
-- No hardcode, no secrets in logs/errors
-- Secrets live in env/secret manager, there is a check for availability
-
-2) Input validation
-- Border schemes (whitelist validation)
-- Bugs don't reveal the insides
-
-3) Injection
-- SQL/NoSQL/command: no concatenations, only parameterization/ORM
-
-4) AuthN/AuthZ
-- Tokens not in localStorage (if applicable), preferably httpOnly cookies
-- AuthZ checks before critical operations
-- RLS/data policies (if you use Supabase/analogue)
-
-5) XSS/CSP
-- Sanitize custom HTML (if any)
-- CSP/headers if required by the architecture
-
-6) CSRF (if cookie-based sessions)
-- CSRF tokens / SameSite=Strict, double-submit pattern if necessary
-
-7) Rate limiting
-- Basic limits on API, stricter for expensive endpoints
-
-8) Sensitive data exposure
-- No PII/secrets in logs
-- No stack trace for user
-
-9) Dependency security
-- audit/lockfile/reproducible install
-
-## Exit
-- Findings P0/P1/P2 + specific fixes
-- What to add to security tests (auth/authz/validation/rate limiting)
-
-## See also
-- Examples and anti-examples: $review_reference_snippets
+---
+name: security_review
+description: AppSec review for endpoints/authorization/input/secrets/sensitive features. Checklist + patterns (OWASP baseline).
+---
+
+# Skill: Security Review (AppSec)
+
+Deep review security code: from secrets to injection.
+
+**Sections:**
+1. [When to activate](#1-when)
+2. [Checklist](#2-checklist)
+3. [DO / DON'T Examples](#3-examples)
+4. [Automated Scans](#4-automated)
+5. [Output Template](#5-output)
+
+---
+
+## 1. When to activate
+
+| Trigger | Depth |
+|---------|-------|
+| New API endpoint | Full review |
+| AuthN/AuthZ changes | Focus: auth section |
+| User input / file uploads | Focus: injection, XSS |
+| Secrets/credentials touched | Focus: secrets |
+| Payments / PII / sensitive data | Full review + GDPR check |
+| Third-party integration | Focus: SSRF, secrets, injection |
+| Webhook handler | Focus: auth, injection, idempotency |
+
+---
+
+## 2. Checklist
+
+### 2.1 Secrets Management
+
+| # | Check | Severity | Status |
+|---|-------|----------|--------|
+| SEC-01 | No hardcoded secrets (passwords, API keys, tokens) | 🔴 P0 | ☐ |
+| SEC-02 | Secrets in env vars only, `.env.example` committed (no values) | 🔴 P0 | ☐ |
+| SEC-03 | pino redact configured for `password`, `token`, `authorization`, `cookie` | 🟠 P1 | ☐ |
+| SEC-04 | `.gitignore` includes `.env`, `*.pem`, `*.key` | 🔴 P0 | ☐ |
+| SEC-05 | Secrets validated at startup (fail fast if missing) | 🟠 P1 | ☐ |
+
+### 2.2 Input Validation
+
+| # | Check | Severity | Status |
+|---|-------|----------|--------|
+| SEC-06 | Zod schema on every endpoint (whitelist approach) | 🔴 P0 | ☐ |
+| SEC-07 | Validation in middleware, before business logic | 🟠 P1 | ☐ |
+| SEC-08 | Error response doesn't expose internals (no Mongoose, no stack) | 🔴 P0 | ☐ |
+| SEC-09 | Numeric/string limits enforced (max length, min/max value) | 🟠 P1 | ☐ |
+
+### 2.3 Injection
+
+| # | Check | Severity | Status |
+|---|-------|----------|--------|
+| SEC-10 | No string concatenation for queries (`Model.find(req.query)` = P0) | 🔴 P0 | ☐ |
+| SEC-11 | Safe filter builder for MongoDB (no `$where`, no user `$` operators) | 🔴 P0 | ☐ |
+| SEC-12 | No `eval()`, `Function()`, `child_process.exec(userInput)` | 🔴 P0 | ☐ |
+| SEC-13 | Mongoose `strictQuery: true` | 🟠 P1 | ☐ |
+
+### 2.4 Auth & AuthZ
+
+| # | Check | Severity | Status |
+|---|-------|----------|--------|
+| SEC-14 | Auth check on every protected endpoint | 🔴 P0 | ☐ |
+| SEC-15 | IDOR prevention: verify ownership before data access | 🔴 P0 | ☐ |
+| SEC-16 | Tokens not in localStorage (prefer httpOnly cookies or server-side) | 🟠 P1 | ☐ |
+| SEC-17 | JWT expiration enforced, refresh flow defined | 🟠 P1 | ☐ |
+
+### 2.5 XSS & CSP
+
+| # | Check | Severity | Status |
+|---|-------|----------|--------|
+| SEC-18 | No `dangerouslySetInnerHTML` without DOMPurify | 🔴 P0 | ☐ |
+| SEC-19 | User-generated content sanitized before rendering | 🔴 P0 | ☐ |
+| SEC-20 | Helmet configured with CSP (if applicable) | 🟠 P1 | ☐ |
+
+### 2.6 CSRF
+
+| # | Check | Severity | Status |
+|---|-------|----------|--------|
+| SEC-21 | Cookie-based auth: `SameSite=Strict/Lax` | 🟠 P1 | ☐ |
+| SEC-22 | CSRF token for state-changing operations (if cookie auth) | 🟠 P1 | ☐ |
+
+### 2.7 Rate Limiting
+
+| # | Check | Severity | Status |
+|---|-------|----------|--------|
+| SEC-23 | Rate limiting on public endpoints | 🟠 P1 | ☐ |
+| SEC-24 | Stricter limits on auth endpoints (login, register) | 🟠 P1 | ☐ |
+| SEC-25 | Rate limiting on expensive operations (file upload, search) | 🟡 P2 | ☐ |
+
+### 2.8 Data Exposure
+
+| # | Check | Severity | Status |
+|---|-------|----------|--------|
+| SEC-26 | No stack trace in production error responses | 🔴 P0 | ☐ |
+| SEC-27 | No PII/secrets in logs | 🔴 P0 | ☐ |
+| SEC-28 | `.select()` on queries — don't return unnecessary fields | 🟠 P1 | ☐ |
+| SEC-29 | Response doesn't include internal IDs unnecessarily | 🟡 P2 | ☐ |
+
+---
+
+## 3. DO / DON'T Examples
+
+### Input Validation
+
+```javascript
+// ❌ DON'T: No validation, raw body to DB
+router.post('/coupons', async (req, res) => {
+  const coupon = await Coupon.create(req.body); // NoSQL injection risk!
+  res.json(coupon);
+});
+
+// ✅ DO: Zod validation middleware, safe create
+router.post('/coupons', validate(createCouponSchema), async (req, res) => {
+  const coupon = await couponService.create(req.validated);
+  res.status(201).json({ data: coupon });
+});
+```
+
+### NoSQL Injection
+
+```javascript
+// ❌ DON'T: User input directly in query
+const user = await User.findOne({ email: req.query.email }); // $gt injection
+// Attack: ?email[$gt]= → matches ALL documents
+
+// ✅ DO: Safe filter builder
+const email = String(req.query.email); // Force string, strip operators
+const user = await User.findOne({ email });
+```
+
+### IDOR (Insecure Direct Object Reference)
+
+```javascript
+// ❌ DON'T: No ownership check
+router.get('/settings/:id', async (req, res) => {
+  const settings = await Settings.findById(req.params.id);
+  res.json(settings); // Any user can read ANY settings!
+});
+
+// ✅ DO: Verify ownership
+router.get('/settings/:id', auth, async (req, res) => {
+  const settings = await Settings.findOne({
+    _id: req.params.id,
+    appInstanceId: req.appInstanceId  // Only own data
+  });
+  if (!settings) return res.status(404).json({ error: { code: 'NOT_FOUND' } });
+  res.json({ data: settings });
+});
+```
+
+### Error Exposure
+
+```javascript
+// ❌ DON'T: Expose internals
+app.use((err, req, res, next) => {
+  res.status(500).json({
+    error: err.message,    // Mongoose error with field names
+    stack: err.stack,      // Full stack trace!
+    query: err.query       // DB query details!
+  });
+});
+
+// ✅ DO: Safe error handler
+app.use((err, req, res, next) => {
+  req.log.error({ err, requestId: req.id }, 'Unhandled error');
+  res.status(500).json({
+    error: { code: 'INTERNAL_ERROR', message: 'Something went wrong' }
+  });
+});
+```
+
+### Secrets in Code
+
+```javascript
+// ❌ DON'T
+const API_KEY = 'sk-live-abc123xyz';  // Hardcoded!
+console.log('Token:', user.refreshToken); // PII in logs!
+
+// ✅ DO
+const API_KEY = process.env.WIX_APP_SECRET;
+if (!API_KEY) throw new Error('WIX_APP_SECRET is required');
+logger.info({ userId: user.id }, 'User authenticated'); // No token
+```
+
+---
+
+## 4. Automated Scans
+
+### Grep patterns
+
+```bash
+# 🔴 P0 — Immediate action
+grep_search: Query="eval("                         → Code injection
+grep_search: Query="dangerouslySetInnerHTML"        → XSS risk
+grep_search: Query="password.*=.*['\"]"             → Hardcoded secret
+grep_search: Query="apiKey.*=.*['\"]"               → Hardcoded secret
+grep_search: Query="Model.find(req."               → NoSQL injection
+grep_search: Query="child_process"                  → Command injection risk
+grep_search: Query="\\$where"                       → MongoDB $where injection
+
+# 🟠 P1 — Review needed
+grep_search: Query="console.log"                    → Replace with pino
+grep_search: Query="localStorage.setItem.*token"    → Token in storage
+grep_search: Query="res.status(500).json.*err"      → Error exposure
+grep_search: Query="cors({ origin: '*'"             → Open CORS
+```
+
+### CLI commands
+
+```bash
+# Dependency audit
+npm audit --audit-level=high
+
+# Secret scanning
+npx secretlint "**/*"
+
+# Outdated packages
+npm outdated
+```
+
+---
+
+## 5. Output Template
+
+```markdown
+# Security Review Report
+
+**Scope:** <PR/feature/module>
+**Reviewer:** Reviewer Agent
+**Date:** YYYY-MM-DD
+**OWASP categories checked:** 8/10
+
+## Findings
+
+| # | Severity | Category | File:Line | Finding | Fix |
+|---|----------|----------|-----------|---------|-----|
+| 1 | 🔴 P0 | Injection | `routes/settings.js:32` | Raw `req.body` passed to `Model.create()` | Add Zod validation middleware |
+| 2 | 🔴 P0 | IDOR | `controllers/coupon.js:18` | No ownership check on DELETE | Verify `appInstanceId` match |
+| 3 | 🟠 P1 | Headers | `server.js` | No Helmet middleware | Add `app.use(helmet())` |
+
+## Automated Scan Results
+| Scan | Result |
+|------|--------|
+| npm audit | ✅ No high/critical |
+| secretlint | ✅ Clean |
+| grep: eval/exec | ✅ Clean |
+| grep: hardcoded secrets | ⚠️ 1 finding (SEC-01) |
+
+## Checklist Summary
+- P0: X findings (must fix)
+- P1: Y findings (should fix)
+- P2: Z findings (optional)
+
+## Security Test Recommendations
+- [ ] Test: invalid input → 400 (not 500)
+- [ ] Test: wrong appInstanceId → 403
+- [ ] Test: duplicate create → 409
+- [ ] Test: missing auth → 401
+
+## Verdict
+- ✅ SECURE — no P0/P1
+- ⚠️ CONDITIONAL — fix P0 before merge
+- ❌ BLOCKED — critical vulnerabilities found
+```
+
+---
+
+## See also
+- `$security_baseline_dev` — implementation patterns (Zod, Helmet, safe filter)
+- `$threat_model_baseline` — architecture-level threat model
+- `$code_review_checklist` — general code review
+- `$review_reference_snippets` — more DO/DON'T examples

package/locales/en/.agents/security_review_baseline/SKILL.md

@@ -1,25 +1,119 @@
 ---
 name: security_review_baseline
-description: Security review based on baseline: OWASP risks, authz, validation, SSRF/XSS/CSRF, secrets, secure logs/errors, least privilege.
+description: Security review by baseline — OWASP-risks, authz, validation, SSRF/XSS/CSRF, secrets, safe logs/errors, least privilege.
 ---
 
 # Skill: Security Review Baseline
 
-## Check (minimum)
-- Input validation on boundaries (API/forms)
-- AuthN/AuthZ: server checks rights, client is not trusted
-- Errors: no leak stacktrace/SQL/configs/PII
-- Secrets: no keys/tokens in repo and logs
-- SSRF: if there is a fetch of external URLs → allowlist/block of internal addresses
-- XSS: escaping/sanitizing (if there is custom HTML)
-- CSRF: if cookie-based sessions → protection (tokens/sameSite)
-- Rate limiting / brute force for sensitive endpoints (if needed)
-- Audit trail for critical operations (if threat model is required)
-
-## Exit
-- P0 security blockers
-- P1 security issues
-- Recommended measures (specifically)
+Quick security-review by the baseline checklist. For deep review → `$security_review`.
+
+> **When use:** Quick baseline with each review (5-10 min).
+> **For deep review** specific areas → `$security_review`.
+
+**Sections:**
+1. [Quick Checklist](#1-checklist)
+2. [Automated Scan](#2-scan)
+3. [Quick Reference](#3-reference)
+4. [Output Template](#4-output)
+
+---
+
+## 1. Quick Checklist
+
+| # | Category | Check | Severity | Status |
+|---|----------|-------|----------|--------|
+| SB-01 | **Secrets** | No hardcoded passwords/tokens/API keys in code | 🔴 P0 | ☐ |
+| SB-02 | **Secrets** | No secrets in logs or error responses | 🔴 P0 | ☐ |
+| SB-03 | **Secrets** | `.env` in `.gitignore`, `.env.example` committed | 🔴 P0 | ☐ |
+| SB-04 | **Validation** | Zod/schema validation on every API endpoint | 🔴 P0 | ☐ |
+| SB-05 | **Validation** | Error responses don't expose internals (no stack, no DB errors) | 🔴 P0 | ☐ |
+| SB-06 | **Auth** | Server-side auth check on protected endpoints | 🔴 P0 | ☐ |
+| SB-07 | **Auth** | IDOR prevention (ownership verification before data access) | 🔴 P0 | ☐ |
+| SB-08 | **Injection** | No `eval()`, `Function()`, `child_process.exec(userInput)` | 🔴 P0 | ☐ |
+| SB-09 | **Injection** | No raw user input in DB queries (NoSQL injection) | 🔴 P0 | ☐ |
+| SB-10 | **XSS** | No `dangerouslySetInnerHTML` without sanitize | 🔴 P0 | ☐ |
+| SB-11 | **CSRF** | Cookie auth → `SameSite=Strict/Lax` + CSRF token | 🟠 P1 | ☐ |
+| SB-12 | **SSRF** | User-controlled URL fetch → allowlist + block internal ranges | 🔴 P0 | ☐ |
+| SB-13 | **Rate Limit** | Rate limiting on public/expensive endpoints | 🟠 P1 | ☐ |
+| SB-14 | **Logging** | No PII in logs (email masked, no passwords/tokens) | 🔴 P0 | ☐ |
+| SB-15 | **Logging** | Audit trail for critical operations | 🟠 P1 | ☐ |
+| SB-16 | **Deps** | `npm audit` — no high/critical vulnerabilities | 🟠 P1 | ☐ |
+| SB-17 | **Headers** | Helmet configured (or security headers set) | 🟠 P1 | ☐ |
+
+---
+
+## 2. Automated Scan
+
+Run these during every review:
+
+```bash
+# 🔴 P0 checks
+grep_search: Query="eval("                          → Code injection
+grep_search: Query="dangerouslySetInnerHTML"         → XSS risk
+grep_search: Query="password.*=.*['\"]"              → Hardcoded secret
+grep_search: Query="apiKey.*=.*['\"]"                → Hardcoded secret
+grep_search: Query="token.*=.*['\"]"                 → Hardcoded secret
+grep_search: Query="Model.find(req."                → NoSQL injection
+grep_search: Query="child_process"                   → Command injection
+grep_search: Query="\\$where"                        → MongoDB injection
+
+# 🟠 P1 checks
+grep_search: Query="console.log"                     → Replace with pino
+grep_search: Query="cors({ origin: '*'"              → Open CORS
+grep_search: Query="eslint-disable"                  → Rule bypass
+
+# CLI
+npm audit --audit-level=high
+```
+
+---
+
+## 3. Quick Reference
+
+For DO/DON'T examples → see `$review_reference_snippets`:
+- Secrets: sections A, B
+- Validation: section C
+- Errors: section D
+- Injection: sections E, E2
+- Auth: section G
+- XSS: section I
+- SSRF: section J
+- CSRF: section H
+
+---
+
+## 4. Output Template
+
+```markdown
+# Security Baseline Check
+
+**Scope:** <PR/module>
+**Date:** YYYY-MM-DD
+
+## Quick Scan Results
+| Check | Result |
+|-------|--------|
+| Hardcoded secrets | ✅ Clean |
+| eval/exec | ✅ Clean |
+| npm audit | ⚠️ 1 moderate |
+| XSS risk | ✅ Clean |
+
+## Findings
+| # | Severity | Check | File:Line | Finding | Fix |
+|---|----------|-------|-----------|---------|-----|
+| 1 | 🔴 P0 | SB-07 | `routes/settings.js:25` | No ownership check | Add appInstanceId verification |
+
+## Summary
+- P0: X findings → must fix
+- P1: Y findings → should fix
+
+## Need deeper review?
+→ Activate `$security_review` for: [auth / injection / SSRF / etc.]
+```
+
+---
 
 ## See also
-- Examples and anti-examples: $review_reference_snippets
+- `$security_review` — deep AppSec review (29 checks)
+- `$review_reference_snippets` — DO/DON'T code examples (A-V)
+- `$threat_model_baseline` — architecture-level threats

package/locales/en/.agents/security_review_baseline/agents/claude.json

@@ -1,12 +1,12 @@
 {
   "name": "security_review_baseline",
   "display_name": "Security Review Baseline",
-  "description": "Security review based on baseline: OWASP risks, authz, validation, SSRF/XSS/CSRF, secrets, secure logs/errors, least privilege.",
+  "description": "Security review by baseline — OWASP-risks, authz, validation, SSRF/XSS/CSRF, secrets, safe logs/errors, least privilege.",
   "default_prompt": "Use $security_review_baseline when the task matches the \"Security Review Baseline\" skill.",
   "triggers": [
     "security_review_baseline",
     "security review baseline",
-    "Security review based on baseline"
+    "Security review by baseline — OWASP-risks, authz, validation, SSRF/XSS/CSRF, secrets, safe logs/errors, least privilege"
   ],
   "capabilities": [
     "security",

package/locales/en/.agents/security_review_baseline/agents/copilot.json

@@ -1,12 +1,12 @@
 {
   "name": "security_review_baseline",
   "display_name": "Security Review Baseline",
-  "description": "Security review based on baseline: OWASP risks, authz, validation, SSRF/XSS/CSRF, secrets, secure logs/errors, least privilege.",
+  "description": "Security review by baseline — OWASP-risks, authz, validation, SSRF/XSS/CSRF, secrets, safe logs/errors, least privilege.",
   "default_prompt": "Use $security_review_baseline when the task matches the \"Security Review Baseline\" skill.",
   "triggers": [
     "security_review_baseline",
     "security review baseline",
-    "Security review based on baseline"
+    "Security review by baseline — OWASP-risks, authz, validation, SSRF/XSS/CSRF, secrets, safe logs/errors, least privilege"
   ],
   "capabilities": [
     "security",

package/locales/en/.agents/security_review_baseline/agents/gemini.json

@@ -1,12 +1,12 @@
 {
   "name": "security_review_baseline",
   "display_name": "Security Review Baseline",
-  "description": "Security review based on baseline: OWASP risks, authz, validation, SSRF/XSS/CSRF, secrets, secure logs/errors, least privilege.",
+  "description": "Security review by baseline — OWASP-risks, authz, validation, SSRF/XSS/CSRF, secrets, safe logs/errors, least privilege.",
   "default_prompt": "Use $security_review_baseline when the task matches the \"Security Review Baseline\" skill.",
   "triggers": [
     "security_review_baseline",
     "security review baseline",
-    "Security review based on baseline"
+    "Security review by baseline — OWASP-risks, authz, validation, SSRF/XSS/CSRF, secrets, safe logs/errors, least privilege"
   ],
   "capabilities": [
     "security",

package/locales/en/.agents/security_review_baseline/agents/openai.yaml

@@ -1,6 +1,6 @@
 interface:
   display_name: "Security Review Baseline"
-  short_description: "Security review based on baseline: OWASP risks, authz, validation, SSRF/XSS/CSRF, secrets, secu…"
+  short_description: "Security review by baseline — OWASP-risks, authz, validation, SSRF/XSS/CSRF, secrets, safe logs…"
   default_prompt: "Use $security_review_baseline when the task matches the \"Security Review Baseline\" skill."
 dependencies:
   tools: []

package/locales/en/.agents/security_review_baseline/agents/qwen.json

@@ -1,12 +1,12 @@
 {
   "name": "security_review_baseline",
   "display_name": "Security Review Baseline",
-  "description": "Security review based on baseline: OWASP risks, authz, validation, SSRF/XSS/CSRF, secrets, secure logs/errors, least privilege.",
+  "description": "Security review by baseline — OWASP-risks, authz, validation, SSRF/XSS/CSRF, secrets, safe logs/errors, least privilege.",
   "default_prompt": "Use $security_review_baseline when the task matches the \"Security Review Baseline\" skill.",
   "triggers": [
     "security_review_baseline",
     "security review baseline",
-    "Security review based on baseline"
+    "Security review by baseline — OWASP-risks, authz, validation, SSRF/XSS/CSRF, secrets, safe logs/errors, least privilege"
   ],
   "capabilities": [
     "security",

package/locales/en/.agents/security_review_baseline/agents/skill.yaml

@@ -1,12 +1,12 @@
 version: 1
 name: "security_review_baseline"
 display_name: "Security Review Baseline"
-description: "Security review based on baseline: OWASP risks, authz, validation, SSRF/XSS/CSRF, secrets, secure logs/errors, least privilege."
+description: "Security review by baseline — OWASP-risks, authz, validation, SSRF/XSS/CSRF, secrets, safe logs/errors, least privilege."
 default_prompt: "Use $security_review_baseline when the task matches the \"Security Review Baseline\" skill."
 triggers:
   - "security_review_baseline"
   - "security review baseline"
-  - "Security review based on baseline"
+  - "Security review by baseline — OWASP-risks, authz, validation, SSRF/XSS/CSRF, secrets, safe logs/errors, least privilege"
 capabilities:
   - "security"
   - "review"