Data / Repositories / Agent security
Agent security repositories on GitHub: 10 ranked by stars
10 public repositories tagged Agent security are on GitHub, with 39,798 stars between them; 10 had a push in the last 30 days. Most are written in Python (8), Rust (1), Go (1). The most starred: NVIDIA/SkillSpector (20,177), Nasiko-Labs/nasiko (9,490), Tencent/AI-Infra-Guard (6,868).
SpecStream is the live web data engine for teams and agents. It reads open-source repositories on GitHub and keeps what it reads, and this page groups that reading by topic. Ask for the repositories you track by topic, language, license or activity, and new ones and changes arrive in your sheet, CRM, Slack or agent.
Every repository, most starred first
| Repository | What it is | Language | Stars | Forks | License | Last push | Read |
|---|---|---|---|---|---|---|---|
| NVIDIA/SkillSpector | Security scanner for AI agent skills. Detect vulnerabilities, malicious patterns, security risks, prompt injection, data exfiltration, and supply-chain risks in Claude Code… | Python | 20,177 | 1,776 | Apache-2.0 | Oct 14 | Oct 14 |
| Nasiko-Labs/nasiko | The Open Runtime for AI Agents | Rust | 9,490 | 2,053 | Oct 7 | Oct 14 | |
| Tencent/AI-Infra-Guard | A full-stack AI Red Teaming platform securing AI ecosystems via Agent Scan, Skills Scan, MCP scan, AI Infra scan and LLM jailbreak evaluation. | Python | 6,868 | 661 | Apache-2.0 | Oct 14 | Oct 14 |
| uber/ADR | ADR secures enterprise AI agents through observability, security benchmarking, and threat detection. Deployed at Uber. | Python | 2,040 | 219 | Apache-2.0 | Oct 13 | Oct 14 |
| hashgraph-online/hol-guard | Open-source antivirus for AI agents: block risky tools, secret access, prompt injection, malicious packages, MCP servers, plugins, and skills at runtime. | Python | 892 | 158 | Apache-2.0 | Oct 14 | Oct 14 |
| graygnatconsole/mcp-audit-tool | Security audit CLI for Model Context Protocol (MCP) servers — scan AI agent configs for tool poisoning, rug pulls, hardcoded secrets, command injection & supply-chain risks. Pure… | Python | 306 | 5 | MIT | Sep 26 | Oct 15 |
| safe-agentic-world/nomos | Deny-wins policy hook for Claude Code and Codex. Checks every shell, file, and fetch call against rules you keep in Git, holds under --dangerously-skip-permissions, refuses… | Go | 19 | 1 | Apache-2.0 | Sep 27 | Oct 14 |
| zrk222/code-factory | Evidence-first audits for AI-generated code: prove tests can fail, catch security and release gaps, and ship reviewable receipts across VS Code, JetBrains, CI, MCP, and App Store… | Python | 6 | 0 | Oct 14 | Oct 14 | |
| basitalisandhu/agent-skills-state-2026 | Study of 2,999 public agent skill, plugin and MCP repositories (October 2026): missing security policies, unpinned MCP servers, unrestricted shell in skills, front matter that… | Python | 0 | 0 | MIT | Oct 13 | Oct 14 |
| tomelias10/mcp-drift-check | Static MCP package-reference mutability checker: Python CLI and GitHub Action. No discovered server execution. Broader config/baseline tools: Orynval Labs. | Python | 0 | 5 | MIT | Oct 11 | Oct 14 |
By language
- Python 8
Related topics
- AI security 8
- MCP 8
- Claude Code 7
- Prompt injection 6
- Supply chain security 5
- AI agents 5
- Model Context Protocol 5
- Agentic AI 4
- LLM security 4
- Cursor 4
- Security scanner 3
- Security tools 3
Read from GitHub by SpecStream between October 14, 2026 and October 15, 2026; each row shows the day its repository was read. Stars and pushes change every day, and the GitHub page is the current one. Forks and archived repositories are left out.