Skillget

awesome-agentic-engineering

mcp
Security Audit
Warn
Health Warn
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 7 GitHub stars
Code Pass
  • Code scan — Scanned 4 files during light audit, no dangerous patterns found
Permissions Pass
  • Permissions — No dangerous permissions requested

No AI report is available for this listing yet.

SUMMARY

Opinionated guide to AI coding agents, workflows, MCP, verification, and safe execution.

README.md

Awesome AI Coding Agent Engineering Awesome

An opinionated, reviewed guide to building reliable AI coding agents. It
focuses on repository rules, coding workflows, tool connections, verification,
and safe execution instead of trying to list every AI agent resource.

Last reviewed: October 8, 2026. Reviews are manual; each review checks the
links and descriptions against their current sources.

Contents

Start Here

Use these six topics as a practical path from agent setup to safe execution
and verification:

Project instructions. Give the agent clear repository rules and durable context.

Coding workflow. Learn a repository-aware working loop from an agent provider.

Permission boundaries. Understand sandboxing, approvals, and command execution.

Tool connections. Expose tools and context through MCP, then inspect them.

Verification. Test the changed system instead of trusting generated code.

Security. Check secrets, dependencies, and execution risk before release.

How This List Is Curated

  • A resource must help build, configure, test, secure, or operate an AI coding
    agent.
  • Official documentation, canonical repositories, and actively maintained
    projects are preferred.
  • Every description must explain a concrete engineering use.
  • Generic AI directories, marketing-only pages, duplicate resources, and stale
    projects are left out.

The goal is a smaller list that helps an engineer make a decision, not a large
collection of links.

Coding Agents

  • Aider - Pairs with language models in the terminal for repository-aware code edits and commits.
  • Claude Code - Runs agentic coding workflows from a terminal with project context and tool use.
  • Cline - Provides an open-source coding agent that can edit files, run commands, and use browser automation.
  • Codex - Provides an open-source coding agent that runs locally in the terminal.
  • GitHub Copilot Cloud Agent - Runs research, planning, and coding tasks autonomously inside GitHub workflows.
  • OpenCode - Offers a terminal coding agent with pluggable model providers and project-aware workflows.
  • Orbi - Turns labelled GitHub issues into pull requests, reviews them in a separate session against the acceptance criteria, then merges and tags a release.
  • SCODE - Provides an open-source terminal coding runtime with persistent sessions, file and shell tools, MCP support, and cloud or local models.
  • yylo - Provides an open-source command-line orchestrator that drives coding agents through typed task, validation, merge, and release-readiness workflows.

Agent Rules And Memories

  • AGENTS.md - Defines a simple project instruction file for coding agents.
  • Claude Code Memory - Shows how Claude Code loads project, user, and local memory files.
  • Coding Agent Guidelines - Packages portable behavioral rules for Claude Code, Cursor, and AGENTS.md consumers.
  • Cursor Rules - Describes project, user, and team rules and AGENTS.md support in Cursor.
  • Evaluating AGENTS.md - Benchmarks how generated and developer-written repository instructions affect coding-agent task success and inference cost.

Claude Code

Codex

  • Codex AGENTS.md - Shows how repository instructions are discovered and merged.
  • Codex Approvals And Security - Explains Codex approval policies, sandboxing, and network controls.
  • Codex Configuration - Documents provider, sandboxing, MCP, and runtime configuration.
  • Codex MCP - Connects Codex to external tools and data through Model Context Protocol.
  • codex-profiles - Provides a community tool for named CODEX_HOME folders and separate local ChatGPT app data on macOS without copying tokens.
  • Codex Sandbox - Explains local sandboxing, approvals, and command execution controls.
  • Codex Skills - Shows where Codex discovers skills and how explicit and description-based invocation works.

Cursor

  • Cursor CLI - Runs Cursor agent workflows and editor commands from a terminal.
  • Cursor Cloud Agents - Runs parallel agents in isolated cloud machines that can build and test changes.
  • Cursor Documentation - Covers editor setup, agents, rules, and model configuration.
  • Cursor MCP - Connects external tools and data sources to Cursor through MCP.

MCP Servers

  • CurlShot MCP - Lets coding agents screenshot a deployed or preview URL in a real browser to check UI changes over remote MCP.
  • deed.page MCP - Lets coding agents publish generated HTML or a static build to a live <slug>.deed.page URL over remote MCP or one HTTP PUT.
  • FastMCP - Provides a Python framework for building MCP servers and clients.
  • MCP Inspector - Debugs MCP servers through an interactive developer tool.
  • MCP Python SDK - Implements the protocol for Python MCP servers and clients.
  • MCP Servers Repository - Lists reference and community MCP server implementations.
  • MCP TypeScript SDK - Implements the protocol for TypeScript MCP servers and clients.
  • Model Context Protocol - Defines the open protocol for connecting agents to tools and context.
  • Rune - Builds a queryable codebase graph with file-and-line evidence for coding agents through MCP.
  • Tale - Connects coding agents to a self-hosted workspace through authenticated MCP tools for knowledge retrieval and automation authoring, testing, and deployment.

Skills And Plugins

  • Agent Skills - Defines the open SKILL.md format that many coding agents load on demand.
  • Claude Code Skills - Bundles task-specific instructions, scripts, and resources for Claude Code.
  • OpenAI Agents SDK - Provides Python primitives for agents, handoffs, tools, guardrails, and tracing.
  • OpenAI Plugins - Packages skills, MCP servers, and optional UI into plugins for ChatGPT and Codex.
  • Smithery - Indexes MCP servers for discovery, installation, and evaluation.

Code Review Workflows

  • CodeRabbit - Reviews pull requests with AI-assisted summaries and inline suggestions.
  • Danger JS - Automates pull-request checks and review comments from JavaScript or TypeScript.
  • GitHub Pull Request Reviews - Documents review states, comments, and requested changes on GitHub.
  • Reviewdog - Reports linter and analyzer findings directly on pull requests.
  • Sourcegraph Cody - Helps Sourcegraph Enterprise users understand, edit, and review code with code-search context.

Testing And Verification

  • Agent QA - Runs natural-language web and mobile regression tests with persistent memory through CLI and MCP interfaces.
  • agent-guard - Blocks test tampering edits and risky outbound commands through Claude Code, Codex CLI, and Gemini CLI hooks. Maintained by @hahahahahahahahah6.
  • agenttrace - Analyzes local AI coding agent sessions for estimated cost, token use, elapsed time, and slow run details. Maintained by @luoyuctl.
  • Jest - Runs JavaScript tests with snapshots, mocks, and watch mode.
  • Mneme - Checks proposed code changes against declared architectural rules through agent integrations and CI. Maintained by @TheoV823.
  • OpenAI Evals - Provides a framework for measuring model behavior and task performance.
  • OrcaReplay - Records coding-agent runs, replays recorded model responses, and forks from checkpoints to compare models.
  • Playwright - Automates browser testing across Chromium, Firefox, and WebKit.
  • pytest - Runs Python tests with fixtures, parametrization, and plugin support.
  • SWE-bench - Benchmarks coding agents on real-world software engineering tasks.
  • Vitest - Provides a fast Vite-native test runner for JavaScript and TypeScript projects.

Security And Sandboxing

  • Gitleaks - Detects secrets and credentials in Git repositories and files.
  • GitHub Secret Scanning - Detects committed secrets and supported token patterns in repositories.
  • OpenSSF Scorecard - Checks projects for common open-source supply-chain security risks.
  • OWASP Top 10 for LLM Applications - Summarizes common security risks in LLM-powered systems.
  • Semgrep - Finds security and correctness issues with customizable static-analysis rules.
  • Socket - Detects supply-chain risk in open-source dependencies.

Prompt Patterns

Case Studies

Learning Resources

Related Public Projects

  • Agent Cost Guard - Local hook that stops supported cloud-cost commands proposed by coding agents until a person approves.
  • Agent Policy Map - Models documented instruction discovery for the same file across Codex CLI, Claude Code, Cursor IDE, and GitHub Copilot CLI.
  • Awesome Agent Instructions - Patterns and a skill for reviewing and simplifying coding-agent instructions while preserving project constraints.
  • Coding Agent Guidelines - Reusable rules for coding agents working in real repositories.
  • Production Launch Prompts - Review prompts and scorecards for checking software before launch.

Contributing

Contributions should add high-signal public resources with a clear engineering
use case. See CONTRIBUTING.md for quality, safety, and review
rules.

Reviews (0)

No results found