Canonical: https://dokima.net/insights/
Updated: 2026-11-18T23:50:32+00:00

Insights & engineering guides

# Practical guides to AI-assisted code security review.

Straight answers on how staged security review works, how to prevent duplicate AI findings, how to run reviews with local models or isolated Unix accounts, and where continuous repository review fits alongside pull-request checks and penetration tests.

[01AI-built Applications7 minute readHow to security-review an app built with Lovable and SupabaseA practical Lovable and Supabase security checklist: review database access rules, privileged functions, secrets and file storage before launch.**Read →**](https://dokima.net/insights/security-review-lovable-supabase-app/)[02AI-built Applications7 minute readHow to security-review your Replit app before launchReview a Replit app before inviting customers: source access, account isolation, payments, admin routes, secrets and background jobs.**Read →**](https://dokima.net/insights/security-review-replit-app-before-launch/)[03Agent Workflows7 minute readUsing Claude Code or Codex to investigate and fix Dokima findingsConnect Claude Code or Codex to Dokima’s read-only findings API, inspect evidence, propose a focused fix and verify it against your code.**Read →**](https://dokima.net/insights/fix-dokima-findings-with-claude-code-or-codex/)[04Security Tooling6 minute readWhere Dokima fits alongside Semgrep, CodeQL and SnykCombine Dokima with Semgrep, CodeQL and Snyk: keep automated checks, investigate application behaviour and carry findings through to verified fixes.**Read →**](https://dokima.net/insights/dokima-alongside-semgrep-codeql-snyk/)[05Engineering Leadership5 minute readYou've fixed a security issue. What should you check next?Three practical questions and a short checklist to help founders and team leads check a security fix before closing the issue.**Read →**](https://dokima.net/insights/what-to-check-after-a-security-fix/)[06Review Architecture8 minute readWhy one-shot AI prompts fail at repository security reviewWhy asking an AI model to find vulnerabilities in a single prompt produces false positives and shallow observations, and how staged review fixes it.**Read →**](https://dokima.net/insights/why-one-shot-ai-prompts-fail-at-security-review/)[07Finding Governance9 minute readHow to run AI vulnerability reviews without getting buried in duplicatesWhy LLMs report the same vulnerability five different ways across runs, and how deterministic fingerprinting and asymmetric merge checks prevent duplicate findings.**Read →**](https://dokima.net/insights/how-to-prevent-duplicate-ai-vulnerability-findings/)[08Data Handling & Models8 minute readRunning AI code security review with local or self-hosted modelsHow engineering teams with strict source-code boundaries can run Dokima using OpenCode and privately hosted local models so repository contents never leave their infrastructure.**Read →**](https://dokima.net/insights/running-ai-code-security-review-with-local-models/)[09Security Engineering8 minute readPull-request security checks vs full-codebase review: why diffs miss serious bugsWhy reviewing only pull-request diffs leaves cross-file authorisation, trust-boundary and state vulnerabilities undetected, and how continuous codebase review closes the gap.**Read →**](https://dokima.net/insights/pull-request-security-checks-vs-full-codebase-review/)[10Engineering Leadership8 minute readHow small engineering teams can run continuous security review without a dedicated AppSec teamA practical workflow for 5-to-40-person engineering teams to find, triage and fix source-code vulnerabilities without hiring a full-time application security specialist.**Read →**](https://dokima.net/insights/continuous-security-review-for-small-engineering-teams/)[11Data Handling & Models8 minute readUsing Claude, Codex or OpenCode for defensive code review: permissions, isolation and riskPractical operational guidance for running AI command-line tools on proprietary repositories, including read-only controls, restricted Unix accounts and Git credential separation.**Read →**](https://dokima.net/insights/claude-codex-opencode-permissions-and-isolation/)[12Data Handling & Models9 minute readClaude vs Codex vs OpenCode for source-code security review: choosing a runner and modelHow Claude Code, OpenAI Codex CLI and OpenCode (including Gemini, open-weights and local models) compare when hunting and verifying software vulnerabilities.**Read →**](https://dokima.net/insights/claude-vs-codex-vs-opencode-for-security-review/)[13Security Engineering9 minute readWhy SAST misses broken access control, IDOR and business-logic flaws (and how AI review finds them)Why pattern-based static analysis struggles to detect missing authorisation checks and tenant isolation bugs, and how staged architectural review catches them.**Read →**](https://dokima.net/insights/why-sast-misses-broken-access-control-and-business-logic/)[14Security Engineering7 minute readSecurity checks before launching code built with Cursor or GitHub CopilotA practical security-review checklist for Cursor and GitHub Copilot users: customer permissions, uploads, secrets, safe configuration and verified fixes.**Read →**](https://dokima.net/insights/auditing-ai-generated-code-for-security-vulnerabilities/)[15Engineering Leadership8 minute readPreparing for a penetration test or SOC 2 / ISO 27001 audit: cleaning up your codebase firstHow engineering teams use staged full-repository security review before external pen tests and compliance audits to eliminate code vulnerabilities early.**Read →**](https://dokima.net/insights/preparing-codebase-for-penetration-test-and-soc2/)[16Review Architecture9 minute readHow to cut false positives in AI code security reviewWhy naive AI security scanners hallucinate vulnerabilities, and how adversarial peer-review stages, upstream context gathering and confidence scoring eliminate noise.**Read →**](https://dokima.net/insights/how-to-cut-false-positives-in-ai-security-review/)[17Review Architecture8 minute readReviewing multi-repository and polyglot architectures for cross-service vulnerabilitiesWhy serious vulnerabilities hide at the boundaries between microservices, frontend apps and backend APIs, and how workspace-level AI security review uncovers them.**Read →**](https://dokima.net/insights/reviewing-multi-repository-and-polyglot-architectures/)[18Cyber Defence Policy7 minute readUsing AI to strengthen cyber defenceWhat recent NCSC and Five Eyes guidance means for engineering teams using AI to improve source-code security review.**Read →**](https://dokima.net/insights/ai-for-cyber-defence/)

See the output directly

## Inspect a real report or run Dokima on your own codebase.

Every guide in this collection reflects how Dokima runs in practice on Linux and macOS: building architectural context first, challenging candidate vulnerabilities before reporting them, and maintaining a deduplicated finding history across commits.

[Start 21-day free trial](https://dokima.net/account/?view=register)[View sample report](https://dokima.net/sample-report/)

See what Dokima finds

## Put an AI security reviewer to work on your codebase.

Start a full-product 21-day trial for your engineering team. No payment card required.

[Start free trial](https://dokima.net/account/?view=register)[Buy Dokima](https://dokima.net/buy/)[View sample report →](https://dokima.net/sample-report/)
