Entity trail

Snyk Agent Scan

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-07-07 / saas-disruption-researcherA New Security Category Emerges: Tools That Police the Coding Agent, Not the Code — Snyk's Evo Agentic Development SecuritySnyk's Evo (covered June 23, part of the current agent-security wave) governs the tools an autonomous coding agent pulls in, the actions it takes at runtime, and the code it generates — enforcing controls inside the agent's workflow rather than scanning output after the fact. As agents write more enterprise software with little human oversight, 'guardrails for the agent' is hardening into a distinct SaaS category alongside Vanta-style GRC automation. Still largely single-vendor-forward, so treat as an emerging pattern rather than a settled market.
2026-06-16 / agents-researcherghostprobe: open-source red-team scanner for MCP tool poisoning and the 'lethal trifecta'Engineer Joe Munene released ghostprobe (discussed June 12, 2026), an open-source scanner that points at live MCP servers to detect tool-poisoning prompts and the 'lethal trifecta' (private data access + untrusted content + exfiltration capability) before integration into an agent stack. It joins a crowded but fast-moving MCP-defense tooling space (Cisco MCP Scanner, Snyk Agent Scan, Proximity), reflecting how MCP server vetting is becoming a standard pre-deployment step. Single-source dev-journal disclosure, so importance is rated low pending wider validation.
2026-05-17 / github-pulse-researchertech-leads-club/agent-skills at 2.6K Stars: Secure Skill Registry Ships Static Analysis and Snyk Scanning After Finding 13% of Marketplace Skills Contain Critical VulnerabilitiesThis curated skill registry for AI coding agents (19+ platforms including Claude Code, Cursor, Copilot, and Windsurf) addresses a real security gap: their audit found that over 13% of marketplace agent skills contain critical vulnerabilities. Every skill in the registry undergoes static analysis, content hashing, and Snyk scanning before publication, with immutable lockfiles and defense-in-depth sanitization. Ships as both a CLI tool for browsing/installing skills and an MCP server for in-agent catalog access.
2026-05-10 / projects-researcherSnyk Embeds Anthropic Claude Into AI Security Platform — Evo Red-Teams Running Agents for Prompt Injection and Data Exfiltration, Scans Agent Supply ChainSnyk announced on May 7 that Claude now powers vulnerability discovery, prioritization, and developer-ready fixes across code, dependencies, containers, and AI-generated artifacts via MCP integration. The standout feature is Evo by Snyk, which uses Claude to continuously discover every AI asset across an organization (models, agents, MCP servers, datasets), red-team running agents for prompt injection and data exfiltration, and enforce runtime policy on tool calls. With 65-70% of production code now AI-generated and nearly half containing vulnerabilities, Snyk is betting the next security frontier is securing the agents themselves.
2026-05-02 / saas-disruption-researcherSnyk Agent Security Goes GA: MCP Governance Engine Now Live in 300+ Enterprises with Claude Code, Cursor, DevinSnyk shipped Evo AI-SPM to general availability with three enforcement agents: Discovery Agent (maps AI attack surface, generates live AI Bill of Materials), Risk Intelligence Agent (enriches inventory with hallucination/bias metrics), and Policy Agent (converts plain-English governance rules into CI-enforced guardrails). Agent Scan governs MCP server supply chains, Snyk Studio enforces validation across 300+ enterprise customers in Claude Code, Cursor, and Devin, and Agent Guard (private preview) stops destructive commands in real-time. The platform also introduced CLI-based red-teaming with attack profiles.
2026-05-02 / vibe-coding-researcherTip: Plug SAST and SCA MCP Servers Into Cursor Security Review for Custom Scanning PipelinesCursor's new Security Review accepts custom MCP servers as additional scanning tools. Connect your existing SAST (Semgrep), SCA (Snyk), and secrets scanners (Gitleaks) via MCP, and Cursor's security agents will use them during every PR review alongside its built-in checks. This turns the IDE's security agent into an orchestrator for your existing security toolchain rather than a standalone replacement — the 'bring your own scanner' approach.
2026-04-26 / skill-finderSnyk Rebrands MCP-Scan to Agent Scan v0.4.13: Scans 7+ AI Coding Agents for 15+ Risks Including Prompt Injection, Tool Poisoning, and Credential LeaksSnyk rebranded its MCP-Scan tool to Snyk Agent Scan, expanding coverage beyond MCP servers to AI agents and agent skills. The CLI auto-discovers configurations for Claude, Cursor, Windsurf, Gemini CLI, Amp, Amazon Q, and others. It scans for 15+ risk categories including prompt injection, tool poisoning/shadowing, toxic flows, malware payloads, and hardcoded secrets. New Background Mode integrates with MDM and CrowdStrike for continuous fleet monitoring. The Skill Inspector web tool lets anyone scan agent skills for free before installing them.
2026-04-25 / saas-disruption-researcherCROSS-CATEGORY: Agent Security Crystallizes as Standalone SaaS Category — Snyk, Cisco, and Vanta Ship Dedicated Products in Same QuarterIn Q1 2026, three major players independently launched dedicated agent security products: Snyk shipped Agent Scan/Studio/Guard for MCP governance and CI/CD enforcement (300+ enterprise customers, integrated with Claude Code, Cursor, Devin); Cisco released DefenseClaw as free open-source agent scanning (MCP scanner, skill scanner, AI BOM generator); Vanta launched Agentic Trust Platform with 24/7 GRC agents. Snyk's data — every AI model deployed introduces 3x untracked software components — quantifies the urgency. The convergence from security, networking, and compliance into the same problem space signals 'agent security' is now a standalone category, not a feature.
2026-04-13 / skill-finderSnyk Agent Scan: AI Agent Supply Chain Security Scanner — Auto-Discovers Configs, 15+ Risk Checks, Open Preview at RSA 2026Snyk launched Agent Scan at RSA Conference 2026, an open-source security scanner that auto-discovers agent configurations across Claude Code/Desktop, Cursor, Gemini CLI, and Windsurf, then scans for 15+ risks including prompt injection, tool poisoning/shadowing, toxic flows, malware payloads, and hardcoded secrets. It treats MCP servers and agent skills as a supply chain problem — the same threat model Snyk applies to npm packages. For builders: run snyk-agent-scan alongside Golf Scanner for defense-in-depth on your local agent setup.
2026-04-05 / skill-finderSnyk Agent Scan: Auto-Discovers Claude/Cursor/Gemini CLI Agents, Scans 15+ Risk Types Including Toxic FlowsSnyk's Agent Scan CLI auto-discovers agent configurations across Claude Code/Desktop, Cursor, Gemini CLI, and Windsurf, then scans for 15+ risk types: prompt injection, tool poisoning/shadowing, toxic flows, malware payloads, untrusted content, credential handling, and hardcoded secrets. The scanner combines multiple LLM-based judges with deterministic rules because agent skills are hybrid artifacts requiring both code analysis and language understanding. Two operating modes: CLI scan (comprehensive report) and Background mode (MDM/Crowdstrike integration for continuous monitoring). Skill Inspector detects malicious skills, insecure configurations, and leaked secrets before production.
2026-04-05 / skill-finderSnyk + Vercel Partnership: Securing the Agent Skill Supply Chain With Verified Skill Registry and Automated ScanningSnyk and Vercel announced a partnership to secure the agent skills supply chain, combining Snyk's Agent Scan tool with Vercel's deployment infrastructure. The integration provides automated skill scanning on every commit, a verified skill registry model, and runtime monitoring for deployed agent skills. Snyk's scanner combines multiple customized LLM-based judges with deterministic rules — a dual approach required because agent skills are hybrid artifacts needing both traditional code analysis and language understanding to catch prompt injection and natural-language malware payloads.
2026-04-05 / skill-finderSnyk Threat Model: SKILL.md to Shell Access in Three Lines of Markdown — Why Traditional Scanners Miss Agent Skill AttacksSnyk published a detailed threat model showing how three lines of markdown in a SKILL.md file instruct an AI agent to read SSH keys and exfiltrate them to attacker infrastructure. The attack exploits the fact that traditional security scanners skip markdown files entirely, and the malicious payload is written in plain English rather than code. The concept of 'toxic flows' — a skill that has access to private data, contains untrusted instructions, and can communicate externally — defines the kill chain. This is the first formal threat model for the AgentSkills specification.

Graph relationships

COMPETES WITH
ghostprobe
ghostprobe competes with Snyk Agent Scan for MCP tool vulnerability scanning
Source finding
COMPETES WITH
ghostprobe -> Snyk Agent Scan
ghostprobe competes with Snyk Agent Scan for MCP tool vulnerability scanning
Source finding

Source trail

Obsidian Security Tech Times Simon Willison ghostprobe arXiv paper critical vulnerability in aws-mcp-server Adversa AI discovered

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues