AGENT SUPPLY CHAIN SECURITY

Security built for vibe coding.

AI agents install packages without human review — so we check every package and its source (registry, URL, publisher, name originality) and hand the agent a deterministic verdict: SAFE / UPGRADE / AVOID — exactly at install time.

npm i -g @soverenia/agent-guard  ·  deterministic verdicts · EU-only hosting
S
4.2%
OF 39K+ AI-AGENT PACKAGES CONTAIN MALICIOUS PATTERNS
650K+
VULNERABILITY & MALWARE RECORDS BEHIND EVERY VERDICT
2 Aug 2026
EU AI ACT STARTS COVERING THE TOOL-CALLING LAYER
100%
EU-ONLY HOSTING & DATA — NO US CLOUD ACT EXPOSURE
THE QUIET GAP

CI scans the code. Nobody guards the install.

Traditional scanners (Snyk, Aikido, Trivy) run in CI — after the code is committed. But the agent's decision happens earlier, on the developer machine, at npm install time. That moment is the attack surface.

Famous Chollima (2026)

North Korean campaign specifically designed to bypass AI coding agents — malicious npm/PyPI packages that read like legitimate docs and pass agent "reasoning". Humans would spot it; the agent installs it in seconds.

attack pattern: agent-assisted supply chain compromise

PromptMink

Malware distributed through package READMEs and metadata containing prompt-injection text aimed at the agent's context window — the agent itself becomes the delivery vector.

prompt injection → agent executes attacker instructions

4.2% of AI-agent packages are malicious

Sigil's 2026 scan of 39,000+ packages built for AI agents found malicious patterns in 4.2%. Typosquats, credential stealers, and install-time payloads — invisible to post-commit scanners.

sigilsec.ai — State of AI Agent Supply Chain Security 2026

Even MCP servers had critical CVEs

Reference MCP servers shipped by major AI labs carried critical vulnerabilities. If the tool layer itself can't be trusted blindly, the verdict layer must be deterministic and signed — not another AI guess.

Aikido AutoTriage = AI heuristics → non-deterministic

One view of every dependency your agent touches.

Verdicts right at the agent's decision point — at install time and in every PR. Plus a provenance check of every source: registry, repository URL, publisher, name originality.

soverenia · agent guard
SCANNED
1 284
SAFE
1 271
UPGRADE
11
AVOID
2
⛔ AVOIDpkg:npm/roblox-creator-docs@100.0.2malware match
⬆ UPGRADEpkg:pypi/jinja2@3.0.0 → 3.1.610 advisories
✓ SAFEpkg:npm/lodash@4.17.210 advisories
✓ VERIFIED SOURCEregistry, repo URL, publisher & name checked at install time

Supported ecosystems.

npm PyPI Maven Go crates.io NuGet RubyGems Composer Hex Pub Conda CocoaPods apk deb rpm

Reachability in your app.

The agent checks whether the vulnerable code is actually reachable in your codebase — per framework, not just per package.

Next.js React Express NestJS Vue Django Flask FastAPI LangChain LangGraph SvelteKit Nuxt Astro Fastify Hono React Router tRPC ASP.NET Core Spring Boot Quarkus Gin Echo Axum Ruby on Rails Laravel Symfony Phoenix Flutter
BEYOND THE PACKAGE CONTENTS

We also check where a package comes from.

Malicious packages often look legitimate — the attack is in the name, the URL or the registry, not the code. We verify provenance at install time:

Registry source — only official registries accepted Repository / homepage URL — dead or redirected links flagged Name originality — typosquats & lookalike (homoglyph) names Publisher & maintainer history — sudden ownership changes License changes — e.g. MIT → proprietary between versions Install scripts — pre/postinstall payloads flagged Lockfile integrity — resolved URL vs. registry hash

A package can be clean code and still arrive through a malicious door — we flag that before the install completes, and the whole lockfile is validated in one call (check_install_plan).

agent-guard · response
// your agent asks before install —
// and gets a verdict back instantly
{ "package": "pkg:pypi/jinja2@3.0.0",
  "verdict": "AVOID",
  "advisories": 10,
  "upgrade_to": "3.1.6" }
✓ SAFE

No known vulnerabilitiesTested against continuously updated vulnerability and malware intelligence. The agent confirms relevance in your code.

⬆ UPGRADE

Vulnerability, but a fix existsWe tell you exactly which version to upgrade to — and the agent checks what it breaks.

⛔ AVOID

Malware, typosquat, active exploitKnown malicious packages, lookalike names, actively exploited vulnerabilities. Do not use.

How it works.

A deterministic pipeline that neither a language model nor a malicious package README can influence.

01 THREAT INTEL PIPELINE

Continuously updated threat intel

Vulnerability and malware databases, merged into one consistent view with full provenance — refreshed automatically, always current.

02 VERDICT ENGINE

Code decides, not a model

Versioned rules compute the verdict. Blue = authoritative engine output. Reproducible and auditable — same input, same verdict, no exceptions.

03 AGENT IN YOUR REPO

Context only it has

The agent queries the verdict via MCP and adds relevance: reachability, upgrade impact. Purple = said by the agent. It never rewrites the verdict to "safe".

NON-NEGOTIABLE PRINCIPLE

The verdict is never made by an LLM.

A language model never touches your security decisions. The verdict is computed by deterministic, versioned rules, and every decision has a full audit trail — reproducible end to end, certifiable for audits.

The AI agent can only lower the relevance of a verdict — "not reachable in our code" — it can never rewrite it to "safe". Prompt injection from a package README never reaches the verdict, because the engine does not read it.

COLOR SEMANTICS = ARCHITECTURE
SAFE · green — verdict
UPGRADE · amber — verdict
AVOID · red — verdict
Engine blue — signed by engine (authoritative)
Purple — said by agent (AI context)
★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★

Sovereign by design.

For the public sector, critical infrastructure and regulated industries — no US CLOUD Act exposure, GDPR-compliant from day one. All hosting and data processing is EU-only.

EU-only data & hosting — GDPR / Schrems II ENISA-anchored intel — EU threat data NIS2 — enforced from 2026 CRA — ready for 2027 EU AI Act — tool layer from 2 Aug 2026 Full audit trail — every decision recorded Deterministic process — certifiable
WHY SOVEREIGNIA

The risks of vibe coding, handled.

Vibe coding means your agent writes code and installs dependencies at machine speed — without a human in the loop. These are the risks that come with it, and what we do about each:

Unverified packages

An agent installs what it is told to — a typosquat or lookalike name (axiοs vs axios) passes "looks right" reasoning in seconds. Soverenia checks every package name for originality before install.

→ name originality check at install time

Injection through docs & metadata

Malicious instructions hidden in a package README or description target the agent's context window — the agent itself becomes the delivery vector. Our engine never reads package content, so injected text can't reach the verdict.

→ verdict computed from metadata only, never from package text

Wrong source

An unofficial registry, a dead or redirected repository, a sudden maintainer change or a license switch between versions — clean code can arrive through a malicious door. We verify provenance: registry, URL, publisher history, lockfile integrity.

→ source & URL checks on every install (see above)

Vulnerable frameworks you never chose

A dependency deep in the tree carries a known CVE — you didn't pick it, your agent did. Soverenia flags it with a verdict and tells you exactly which version fixes it, checked against continuously updated threat intelligence.

→ SAFE / UPGRADE / AVOID per package, with upgrade target

Pricing.

Free distribution via the open-source Agent Guard, paid platform for teams and sovereign enterprise deployments.

FREE
0 €
forever · open-source CLI
  • Agent Guard CLI + MCP server
  • Verdicts for public repositories
  • Continuously updated threat intelligence
  • Community support
Read the docs
ENTERPRISE
Custom
sales-led · NDA · DPA
  • Local-hosted / sovereign deployment (EU-only)
  • Audit export, SLA
  • NIS2 / CRA compliance reports
  • Analyst-verified tier (optional)
Contact us

FAQ.

Short answers to the questions we hear most.

How is this different from Snyk, Aikido or Trivy?

Those tools scan your code or images in CI — after the commit. Soverenia sits at the agent's decision point, at npm install time on the developer machine, and hands the agent a deterministic verdict via MCP. It's not a broader scanner; it's the missing layer exactly where agents act.

Is an LLM involved in the verdict?

No. The verdict is computed by deterministic, versioned rules — reproducible and auditable. The AI agent calling us can only add local context (e.g. "not reachable in our codebase") — it can never rewrite a verdict to SAFE.

Which ecosystems do you support?

npm, PyPI, Maven, Go, crates.io, NuGet and RubyGems — full scope today, with continuously updated vulnerability and malware intelligence behind every verdict.

Where is my data hosted — and do you store my queries?

EU-only hosting, no US CLOUD Act exposure — a natural fit for NIS2 / CRA / EU AI Act compliance. We don't store your queries: this website sets no cookies and runs no analytics or third-party trackers (fonts are self-hosted; the only localStorage entry is your theme preference, which never leaves your browser). The only data kept is basic server access logs (IP, timestamp, requested path) for security and rate limiting, deleted after 30 days.

Can I self-host?

Yes — the Enterprise plan includes a local-hosted, sovereign EU-only deployment with audit export and SLA. The CLI (Agent Guard) is open source and free.

Is there a free tier?

Yes — the Agent Guard CLI and MCP server are free and open source, with verdicts for public repositories included. Paid plans (see Pricing above) add private repos, CI gates, policy profiles and sovereign enterprise deployments.

CONTACT

Talk to us about Enterprise.

Local-hosted sovereign deployment (EU-only), audit export, SLA, NIS2/CRA reporting — tell us about your setup and we'll get back within one business day.