Skip to content

Security: veriswarm/veriswarm-sdk

Security

SECURITY.md

Security Policy

Reporting a Vulnerability

If you discover a security vulnerability in the VeriSwarm SDKs or MCP server, please report it responsibly.

Email: security@veriswarm.ai

Please include:

  • Description of the vulnerability
  • Steps to reproduce
  • Affected package(s) and version(s)
  • Impact assessment

Do NOT open a public GitHub issue for security vulnerabilities.

Response Timeline

  • Acknowledgment: Within 48 hours
  • Initial assessment: Within 5 business days
  • Fix for critical issues: Within 7 days
  • Fix for non-critical issues: Within 30 days

Scope

This policy covers:

  • @veriswarm/sdk (Node.js SDK)
  • veriswarm (Python SDK)
  • veriswarm-mcp (MCP Server)
  • @veriswarm/openclaw-plugin (OpenClaw plugin)
  • @veriswarm/schemas (JSON schemas)
  • veriswarm-action (GitHub Action)
  • LangChain adapter (veriswarm[langchain])

For vulnerabilities in the VeriSwarm platform itself (API, web dashboard, infrastructure), please email security@veriswarm.ai directly.

Security Posture

Security-relevant features of the VeriSwarm platform:

  • Authentication: Argon2id password hashing, ES256 JWT credentials, Ed25519 inter-agent message signing, MFA with TOTP and recovery codes
  • Tenant isolation: per-tenant scoping enforced at the database query level on every endpoint
  • Rate limiting: per-IP and per-tenant limits on all authenticated routes
  • IP allowlists: optional per-tenant CIDR-based allowlist enforced via global middleware
  • Encryption at rest: Fernet/AES-256 for PII tokenization and BYOK API keys
  • Encryption in transit: TLS 1.3 enforced
  • Supply chain: Ed25519-signed agent template exports, SHA-256 pinned Docker base images
  • Defense in depth: hash-chained audit ledger (Vault), cross-model verification for memory poisoning, three-state circuit breakers per LLM provider

Supported Versions

We provide security updates for the latest minor version of each package. Older versions are not supported.

Package Supported
Latest release Yes
Previous minor Best effort
Older versions No

Disclosure

We follow coordinated disclosure. We will credit reporters in release notes unless they prefer anonymity. We do not currently offer a bug bounty program.

Advisory Process

Security advisories are published via GitHub Security Advisories (GHSA) on this repository. We will request a CVE through GitHub's CNA for vulnerabilities with a CVSS v3 score of 7.0 or higher (HIGH/CRITICAL). Advisories include:

  • Affected package versions and fixed version
  • CVSS v3 score and vector string
  • Description of the vulnerability and impact
  • Workarounds where applicable
  • Credit to the reporter (with their permission)

Subscribe to GitHub Security Advisories on this repo to receive notifications.

There aren't any published security advisories