Skip to content
Agentmart
Submit a tool

Best agent skills for MCP and agent security

We found 70 agent skills for MCP and agent security in the catalog; these are the 15 that rank highest. The order comes from our rank, which weighs adoption, freshness, documentation and publisher identity. Nobody can pay for a place on this list. ai-agent-security, ai-security-hardening and indirect-injection-detection lead it. indirect-injection-detection and direct-injection-detection are published by the company that makes them. ai-agent-security comes from the largest repository here (48,428 stars). 9 of the 15 had activity in the last 30 days. Each entry lists why it made the list and links to its install page.

Updated Oct 26, 2026 · Built from catalog data; how we rank and grade

Top 15 of 70

  1. 1
    ai-agent-security

    devops-skills

    Secure AI agents against prompt injection, tool abuse, and data exfiltration with defense-in-depth controls.

    • Largest repository here (48,428 stars)
    • Active in the last 30 days
    • Open source (MIT)

    ★ 48.4KLast active Oct 24, 2026Claude Code, Claude apps, Codex, Cursor +6

  2. 2
    ai-security-hardening

    devops-skills

    Harden AI/LLM deployments against prompt injection, data exfiltration, model theft, and supply chain attacks.

    • Largest repository here (48,428 stars)
    • Active in the last 30 days
    • Open source (MIT)

    ★ 48.4KLast active Oct 24, 2026Claude Code, Claude apps, Codex, Cursor +6

  3. 3
    indirect-injection-detection

    TencentOfficial

    Detect indirect prompt injection (goal hijack).

    • Published by the vendor
    • Repository has 7.1K stars
    • Active in the last 30 days

    ★ 7.1KLast active Oct 22, 2026Claude Code, Claude apps, Codex, Cursor +6

  4. 4
    direct-injection-detection

    TencentOfficial

    Detect direct prompt injection or instruction override via user message (no external content).

    • Published by the vendor
    • Repository has 7.1K stars
    • Active in the last 30 days

    ★ 7.1KLast active Oct 22, 2026Claude Code, Claude apps, Codex, Cursor +6

  5. 5
    hunt-llm

    Encod3d-Sec

    LLM / AI application attack hunting - prompt injection (direct + indirect), excessive agency, insecure output handling, system-prompt + data leakage.

    • Open source (MIT)

    ★ 335Last active Aug 19, 2026Claude Code, Claude apps, Codex, Cursor +6

  6. 6
    pleading-injection-guard

    lawve-ai

    Detects prompt injection hidden in documents from the other side (pleadings, skeletons, bundles, served evidence, opponents' emails and attachments) before an…

    • Active in the last 30 days
    • Ships helper scripts
    • For Python and bash

    ★ 895Last active Oct 2, 2026Claude Code, Claude apps, Codex, Cursor +6

  7. 7
    hunt-mcp

    Encod3d-Sec

    MCP server attack hunting - tool poisoning, indirect prompt injection via tool output, rug-pull updates, cross-tool shadowing…

    • Open source (MIT)

    ★ 335Last active Jul 27, 2026Claude Code, Claude apps, Codex, Cursor +6

  8. 8
    pentest-ai-llm-security

    jd-opensource

    AI/LLM application security testing — prompt injection, jailbreaking, data exfiltration, and insecure output handling per OWASP LLM Top 10.

    • Open source (Apache-2.0)

    ★ 315Last active Jul 2, 2026Claude Code, Claude apps, Codex, Cursor +6

  9. 9
    skill-injection-defense

    lawve-ai

    Audits legal AI skills, prompts, workflows, MCP/tool instructions, and agent packages for prompt injection, malicious instructions, unsafe scripts, suspicious…

    ★ 895Last active Jul 30, 2026Claude Code, Claude apps, Codex, Cursor +6

  10. 10
    prompt-guard

    Orchestra-Research

    Meta's 86M prompt injection and jailbreak detector.

    • Repository has 13.9K stars
    • Open source (MIT)

    ★ 13.9KLast active Jun 16, 2026Claude Code, Claude apps, Codex, Cursor +6

  11. 11
    detecting-ai-model-prompt-injection-attacks

    adriannoes

    Detects prompt injection attacks targeting LLM-based applications using a multi-layered defense combining regex pattern matching for known attack signatures…

    • Active in the last 30 days
    • Open source (Apache-2.0)
    • Ships helper scripts

    ★ 72Last active Oct 24, 2026Claude Code, Claude apps, Codex, Cursor +6

  12. 12
    prompt-guard

    profbernardoj

    Advanced prompt injection defense system for EverClaw with HiveFence network integration.

    • Open source (MIT)
    • Ships helper scripts
    • For Python

    ★ 110Last active May 15, 2026Claude Code, Claude apps, Codex, Cursor +6

  13. 13
    ai-engineering-toolkit

    sickn33

    Overview of six AI engineering workflows: prompt evaluation (8-dimension scoring), context budget planning, RAG pipeline design, agent security audit…

    • Largest repository here (48,428 stars)
    • Active in the last 30 days
    • Open source (MIT)

    ★ 48.4KLast active Oct 24, 2026Claude Code, Claude apps, Codex, Cursor +6

  14. 14
    little-canary

    hermes-labs-ai

    Use when you need to screen inbound untrusted text before it reaches a primary model — detecting prompt injection by its behavioral effect on a sacrificial…

    • Active in the last 30 days
    • Open source (Apache-2.0)

    ★ 41Last active Oct 20, 2026Claude Code, Claude apps, Codex, Cursor +6

  15. 15
    llm-security

    coco-research

    Use for authorized security assessment of LLM applications and AI agents, including prompt injection, tool abuse, RAG exposure, memory poisoning, and model…

    • Active in the last 30 days

    ★ 784Last active Oct 23, 2026Claude Code, Claude apps, Codex, Cursor +6

Related