Zero-Day Resilience: No-Exit WASM Sandboxing (Claude/MCP RCE Protected)

ABS Core™.
Elite AI Governance.

The Sovereign Governance: Deterministic Rules & Compliance Protocol that secures your AI Agents.
Intercepts intent, enforces policy, and generates immutable audit trails.

v10.1.4 — Sovereign Proof Edition
$ npm install @abs-core/sdk
Try Policy Sandbox
Documentation

The Immune System Architecture

11 integrated capabilities — from Intent (CHI) to Consensus (Quorum) — plus three Core Vaccines for runtime defense.

OCS
Offline Cost Sentinel

Pre-flight regulatory and cost governor. Blocks policy-violating actions before any network call — offline, no round-trip required.

AICCP
Change Control

Forces a structured decision cycle before any irreversible write operation — no shortcuts.

WASM
Universal Kernel

High-performance Rust Enforcer (<5ms) running in Edge sandbox.

CHI
Cognitive Host Interface

Mandatory intent introspection before execution. Detects semantic drift, PII, and prompt injection in real time.

CORTEX
Agent Memory

Memory compliance and data state introspection to sanitize long-term memory streams.

LEDGER
Forensic Proof

SHA-256 forensic ledger with Bitcoin or L2 anchoring. Every decision is cryptographically verifiable.

CERTIFIED VAULT
Policy Packs

Pre-built YAML policy packs for HIPAA, SOC2, LGPD, and financial compliance. Accelerates audit preparation.

QUORUM
Human Consensus

Enforce multiple M-of-N signatures for high-risk decisions. Zero-Trust approval chains.

IDENTITY SYNC
Okta & LDAP

Automated authority syncing via Webhooks. Inherit corporate hierarchies into governance.

MCP BRIDGE
Tool Firewall

Intercept every tool call before execution. JSON/YAML Enforcement for AI agents.

POLICY DSL
Declarative Rules

Define policies in JSON/YAML instead of code. 7 built-in system policies block credential theft, path traversal, and exfiltration.

Core Vaccines

Proactive defense system protecting your agent's runtime.

SECRET VAULT
JIT Secret Injection

Prevents key leakage by injecting credentials only at runtime. Your agent never sees the API key.

HEARTBEAT SENTINEL
Dead Man's Switch

Health check and recovery monitoring for autonomous agents. Triggers critical alerts if the agent stops reporting.

SHADOW MODE
Risk Analytics

Run policies in observation mode. Measure risk before enforcing blocks.

Seamless Integration

Works with your existing stack.

LANGCHAIN ADAPTER
Native Support

Drop-in integration via @abs-core/adapter-langchain. Intercepts LLM/Tool start events.

CREW AI
Sentinel SDK

Governance for autonomous Python agents via the abs-sdk. The Magic Proxy covers simple single-agent calls; multi-agent frameworks (CrewAI, LangGraph) require explicit SDK integration.

VERCEL AI SDK
Middleware

Zero-code adoption for Next.js apps via @abs-core/vercel-adapter. Seamless Protocol Translation.

Magic Proxy Adoption (Zero-Code for Simple Agents)
const openai = new OpenAI({ baseURL: "https://api.abscore.app/v1" });

// Single-agent, proxy-compatible frameworks. Multi-agent (CrewAI, LangGraph) requires the @abs-core/sdk.

Technical Glossary

Decoding the Immune System Acronyms.

OCS

Offline Cost Sentinel: Pre-flight regulatory and cost governor. Blocks policy-violating and cost-bearing actions before any network call — offline, no round-trip required.

CHI

Cognitive Host Interface: Intent introspection engine. Verifies reasoning before state changes.

AICCP

AI Change Control Protocol: Forces a structured decision cycle before any irreversible write operation — no shortcuts.

Bond / Slash

Financial Accountability: Agents post collateral against policy compliance. Violations trigger automatic slashing — making risk quantifiable and insurable.

Transparent Pricing

Enterprise governance scaled for everyone.

Community

$0/mo
  • Community WASM Kernel
  • Shadow Mode (Observation)
  • Public L2 Ledger
Read the Docs
Enterprise

Sovereign

Custom
  • On-Prem / TEE Deployment
  • Private Bitcoin Anchoring
  • Decision Quorum (M-of-N)
  • Bond / Slash — Financial Accountability
Contact Sales
ENTERPRISE DEMO

DataShield: Data Leak Prevention (DLP)

See the Magic Proxy blocking an agent trying to read .env or exfiltrate AWS credentials, while allowing secure log reading operations.

Schedule Demo (5 min)
$ ./agent.py --read logs/app.log
✔ ALLOWED: Read 45KB (Safe Pattern)
hash: 8f4b2e1... (ledger block #1042)
$ ./agent.py --cat .env
✖ BLOCKED: Policy Violation (SECRET_ACCESS)
Risk Score: 99/100 (Critical)
Vaccine: Infra_Governor_v2
proof_hash: 16ccfa04... (ledger block #1043)
_
$ abs-cli report --standard iso27001 --last 1h
# ISO 27001 Violation Report
Severity: CRITICAL (A.5.15 Violation)

Analysis: The agent attempted to access sensitive configuration files. Blocked by Infra_Governor_v2.
✦ v10.1.4Sovereign Proof

MCP Governance Bridge

Intercept every tool call before execution. Policy-as-Code for AI agents.

MCP BRIDGE
Tool Call Firewall

Intercepts MCP tool calls with layered policy evaluation, entropy analysis, and immutable audit logging.

POLICY DSL
Declarative Rules

Define policies in JSON instead of code. 7 built-in system policies block credential theft, path traversal, and exfiltration.

AGENT JWT
Cryptographic Identity

HMAC-SHA256 tokens with tool permissions, risk budgets, and bond status. Low-latency verification at the edge.

MCP Bridge API
POST /v1/mcp/tool-call // Evaluate before execute
TypeScript / Next.js SDK
npm install @abs-core/sdk # Production Governance