Category guide · updated 2026-10-31
Best Coding & Dev agents
Botlens lists 2,981 bots and skills in Coding & Dev. Of those, 1,550 are marked Safe, 177 Caution and 1,254 not scanned yet. Risky ones are never listed. Below are 20 Safe picks, starting with the ones whose full code we could read.
The best AI coding agents on Botlens are Safe listings with full code read. For code review, Adversarial Code Reviewer returns a merge verdict on recent changes. Address Github Comments handles PR review comments after you approve. Agy Delegate hands off a bounded task, then you review the diff. Botlens marks 1,466 of 2,987 Coding listings Safe.
Last checked: 2026-11-01 · How Botlens checks agents: a rule-based scan of each listing's public instructions and code, with coverage recorded.
| Agent | Best for | Botlens result | What its listing says | Page |
|---|---|---|---|---|
| Adversarial Code Reviewer | Code review | Safe, full code read | Reviews recent code changes through three hostile reviewer personas and returns a merge verdict | /bot/21146 |
| Address Github Comments | PR comments | Safe, full code read | Addresses GitHub PR review comments with gh CLI after user approval | /bot/21128 |
| Agy Delegate | Task handoff | Safe, full code read | Hands a bounded coding task to the Antigravity CLI, then you review and commit the diff | /bot/21241 |
| Kiwi Hook | Workflow | Safe, full code read | Wakes your bot when your coding agent finishes a turn | /bot/8 |
| Aws Mcp Setup | MCP setup | Safe, full code read | Configures AWS MCP servers for documentation search and API access | /bot/21625 |
How to choose an AI coding agent?
Choose by what it may change. An agent that only reads and comments is low risk. One that writes code, runs commands or pushes to your repo needs more trust, so look for listings that say they wait for your approval.
Then match the task. Review, refactoring, PR chores and environment setup are different jobs, and a narrow tool usually beats a do-everything one.
What is a good AI code review bot?
A good AI code review bot reads your diff and reports, and it does not merge on its own. Adversarial Code Reviewer is a Safe example that returns a verdict for you to act on. Address Github Comments works on review comments but only after you approve. Both keep you in charge of what lands in the codebase.
Give a review bot read access to the repo first. Add write access only if you want it to fix things, and keep pushes manual.
Agent skills vs MCP servers: which do I need?
Use a skill when you need instructions and a workflow. Use an MCP server when your agent needs new tools that connect to a service. A skill is mostly text your agent follows. An MCP server runs code and adds callable tools, so it can reach more and needs a closer look. Read how to check an MCP server before installing.
How to choose a coding agent
- Pick one job. Review, PR chores, delegation or setup.
- Filter to Safe with full code read.
- Check write and shell access. Prefer approval before changes.
- Try it on a branch. Never start on main.
- Pin the version. Re-check after updates.
Let Botlens pick for you
Yes. Botlens is a free catalog of Grok Bot bots, templates and skills that shows each listing's safety status, and it ships a Grok Bot skill that picks them for you. Describe your task and the skill searches the catalog, skips listings flagged Risky, ranks matches by fit and safety, and asks before installing anything. You don't have to test bots by hand.
Botlens lists 2,987 Coding & Dev bots and skills: 1,466 marked Safe and 1,458 not scanned yet. Read How to find the right AI bots and skills automatically to set it up.
What is the best AI code review bot?
On Botlens, Adversarial Code Reviewer is a Safe pick with full code read. It reviews recent changes and returns a merge verdict, leaving the merge to you. If your team works in GitHub pull requests, Address Github Comments is another Safe option that acts on review comments only after you approve.
Are agent skills safer than MCP servers?
Often, but not always. A skill is mostly instructions, while an MCP server runs code and adds tools that can reach outside services. Both run with your permissions. Check either one before installing, and look at what it can read, where it connects and whether it asks before acting.
Top Safe picks
Reusable Grok Bot skill and helper scripts that hand a Codex CLI session to the Codex desktop app through macOS Terminal.app.
Creates and manages Cloud SQL instances for MySQL, PostgreSQL, and SQL Server on Google Cloud.
Designs and implements cloud infrastructure using Terraform, Kubernetes, and CI/CD pipelines.
Build and optimize Unity games with C# scripting and cross-platform deployment. You are a Unity game developer expert with 8+ years of experience buil
Grok Bot cloud-instance fork of mcpserver-grok-plugin with .grok-plugin hooks for TODO, session logging, plan sync, and GraphRAG continuity.
Weekly browser mini-game catalog with autonomous-game-development SKILL.md packs and Day1–7 shipping rules for Bot-driven game a week.
Builds and deploys production-ready serverless applications on AWS using Lambda, API Gateway, DynamoDB, and SAM/CDK.
Supports IT consultants across the full software development lifecycle, from code review to deployment.
Helps game developers debug code, optimize performance, and build player-facing troubleshooting tools.
Create, update, or review Azure infrastructure as code in Terraform using Azure Verified Modules.
Build AWS infrastructure with CDK using TypeScript/Python, verified against live AWS docs.
Setup that lets a lead Grok Bot spin up teammate Bots for Claude Code, Codex, Gemini or local CLIs on your own machines: they branch, code and report back while
AI-assisted debugging toolkit smart debug expert for rapid root cause analysis and fix generation. Use this capability when working on debugging toolkit sm
DeepSeek Harness OAuth for coding subscriptions: SuperGrok / Grok Build, ChatGPT Plus Codex, Kimi Code, and Claude Code without pasting API keys.
Provide principal-level software engineering guidance with focus on engineering excellence, technical leadership, and pragmatic implementation.
Self-hosted MCP server so a Grok Bot can hand coding tasks to Devin and get session links or PRs back without exposing the Devin API key.
Designs and reviews cloud security architecture across AWS, Azure and GCP, with zero trust, least privilege and IaC guardrails.
Provisions and hardens Azure SQL Database, Elastic Pools, and Cosmos DB, with backups, geo-replication, and security.
Sets up CI/CD pipelines, scaffolds infrastructure as code, and manages cloud deployments across AWS, GCP, and Azure.
Grok Build skill that asks Codex CLI (GPT-6 Astra) to explore the repo and write a detailed plan.md for Grok.
FAQ
How many Coding & Dev bots are safe?
1,550 of the 2,981 Coding & Dev listings are marked Safe as of 2026-10-31. “Safe” means our heuristic scan found no obvious threats; it is not a guarantee.
How is the list ordered?
Only Safe listings are shown. They are ordered by how closely their name and description match this category, then by how much of their code we could read. There are no paid placements in this list.
How do I install one?
Open the bot page and use “Copy install prompt”, or install the Botlens skill and ask your Grok Bot for the task.
The free Botlens skill searches this catalog, skips anything flagged Risky and asks you before installing.
Other categories: Research · Design & Video · Writing & Content · Productivity · Finance & Crypto · Sales & CRM · Travel & Life · Learning · Marketing & Growth · Health & Fitness · Social Media · Agent Ops & Automation · Business & Ops · Security & Compliance · Shopping & Deals
Read next: Is it safe to install AI agent skills?