awslabs

Agentic Developer Platform — Security skill for Claude Code

Security community

ADP is the foundation for building and running AI agents across an enterprise.

How to install Agentic Developer Platform

This entry records only its repository, not the path inside it, so there is no exact command to give. Open awslabs/agentic-developer-platform and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Agentic Developer Platform does

**ADP is the foundation for building and running AI agents across an enterprise.** Not one product — a base on which many agentic platforms get built (AI Operations, Cyber Threat, Agentic Analytics, …

Alternatives in Security

T3mp3st — autonomous red teaming platform; multi-agent offensive-security meta-harness 5.7k ★
Skills Audit Report — Date: 2026-02-15 Auditor: Automated Skill Quality Audit Scope: Recently added skills in business-growth/, fina 5.3k ★
Anthropic Cybersecurity Skills — 734+ structured cybersecurity skills for AI agents · MITRE ATT&CK mapped · agentskills.io open standard · Work 3.8k ★

README

ADP — Agentic Developer Platform

Concretely: you bring the agent's job, and ADP handles everything around it — model access, scaling, auth, tenant isolation, memory, tool access, audit, and cost controls. Adding a new agent is a five-file task; the platform provides the rest. It's the layer between "a clever prompt that works on your laptop" and "an agent a whole org can safely use against real systems."

What you get out of the box

**Governed model access** — a multi-tenant Amazon Bedrock proxy with per-org/team/user budgets, rate limits, and full audit. Point Claude Code, Cursor, or any OpenAI-compatible client at one endpoint.
**A team of autonomous coding agents** — mention an agent on a GitHub issue or PR; a pod spins up, does the work, and opens a PR. No runner babysitting. The personas cover the full AI development life cycle (AIDLC):
- @agent-product — gathers requirements, writes user stories + acceptance criteria
- @agent-pm — orchestrates the workflow, decomposes work, coordinates the board
- @agent-architect — designs systems, defines interfaces, produces design docs
- @agent-developer — writes production code + tests, opens the PR
- @agent-reviewer — the quality gate: reviews for correctness/security, blocks on real issues
- @agent-operations — deploys, monitors, and maintains infrastructure
Domains add their own (e.g. **@malware-analysis-agent** in the cyber pack), and new personas are a five-file declaration — no platform changes.
**Code intelligence** — one MCP endpoint giving agents semantic search, code search, wikis, and persistent memory across your codebases.
**A shared harness** — tool routing, jobs, events, ar