# ๐Ÿ›ก๏ธ OWASP Agentic Top 10 โ€” Compliance Mapping **How the Agent Governance stack covers the [OWASP Top 10 for Agentic Applications (2026)](https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026/)**
--- ## Coverage Summary | # | OWASP Risk | Coverage | Component | |---|-----------|----------|-----------| | ASI-01 | Agent Goal Hijack | โœ… Covered | Agent OS โ€” Policy Engine | | ASI-02 | Tool Misuse & Exploitation | โœ… Covered | Agent OS โ€” Capability Sandboxing | | ASI-03 | Identity & Privilege Abuse | โœ… Covered | AgentMesh โ€” DID Identity & Trust Scoring | | ASI-04 | Agentic Supply Chain Vulnerabilities | โœ… Covered | AgentMesh โ€” AI-BOM (model + data + weights provenance) | | ASI-05 | Unexpected Code Execution | โœ… Covered | Agent Runtime โ€” Execution Rings | | ASI-06 | Memory & Context Poisoning | โœ… Covered | Agent OS โ€” VFS Policies + CMVK Verification | | ASI-07 | Insecure Inter-Agent Communication | โœ… Covered | AgentMesh โ€” IATP + Encrypted Channels | | ASI-08 | Cascading Failures | โœ… Covered | Agent SRE โ€” Circuit Breakers + SLOs | | ASI-09 | Human-Agent Trust Exploitation | โœ… Covered | Agent OS โ€” Approval Workflows | | ASI-10 | Rogue Agents | โœ… Covered | Agent Runtime โ€” Kill Switch + Ring Isolation | **Mappings in place for all 10 risk categories.** Coverage is provided through the combined governance stack; deployers should pair these controls with the layered defenses described in `docs/LIMITATIONS.md` for production use. --- ## Detailed Mapping ### ASI-01: Agent Goal Hijack > *Attackers manipulate the agent's objectives via indirect prompt injection or poisoned inputs.* **Mitigation:** Agent OS enforces **policy-based action interception** at the application layer. Every agent action passes through the policy engine before execution. Unauthorized goal changes are blocked before they reach the agent's tools. - **Policy Engine** โ€” declarative rules controlling what agents can and cannot do - **Action Interception** โ€” governance middleware intercepts agent actions before execution - **Policy Modes** โ€” `strict` (deny by default), `permissive` (allow by default), `audit` (log only) - **MCP Governance Proxy** โ€” policy enforcement for MCP tool calls ```python from agent_os import StatelessKernel, ExecutionContext kernel = StatelessKernel() ctx = ExecutionContext(agent_id="my-agent", policies=["read_only"]) # This action is blocked by policy โ€” goal hijack prevented result = await kernel.execute( action="delete_database", params={"target": "production"}, context=ctx, ) # result.success = False, result.error = "Policy violation: read_only" ``` **Component:** [Agent OS](https://github.com/microsoft/agent-governance-toolkit) โ€” `src/agent_os/policy/`, `extensions/mcp-server/src/services/policy-engine.ts` --- ### ASI-02: Tool Misuse & Exploitation > *An agent's authorized tools are abused in unintended ways, such as exfiltrating data via read operations.* **Mitigation:** Agent OS provides **capability-based security** inspired by POSIX. Agents are granted specific, scoped capabilities โ€” not blanket tool access. Tool inputs are sanitized for injection patterns. - **Capability Sandboxing** โ€” agents receive explicit capability grants (read, write, execute, network) - **Tool Allowlists/Denylists** โ€” built-in strict mode blocks `run_shell`, `execute_command`, `eval` - **Input Sanitization** โ€” command injection detection, shell metacharacter blocking - **`verify_code_safety`** MCP tool โ€” checks generated code before execution **Component:** [Agent OS](https://github.com/microsoft/agent-governance-toolkit) โ€” capability model, MCP proxy policy rules --- ### ASI-03: Identity & Privilege Abuse > *Agents escalate privileges by abusing identities or inheriting excessive credentials.* **Mitigation:** AgentMesh implements **zero-trust identity** using Decentralized Identifiers (DIDs). Every agent has a cryptographic identity with scoped capabilities. Trust is earned, not assumed. - **DID Identity** โ€” `did:agentmesh:{agentId}:{fingerprint}` with Ed25519 key pairs - **Trust Scoring** โ€” tiered model: `Untrusted โ†’ Provisional โ†’ Trusted โ†’ Verified` - **Delegation Chains** โ€” track trust inheritance with verifiable credentials - **Challenge-Response Handshake** โ€” cryptographic authentication at connection time - **Trust Decay** โ€” scores degrade over time without positive signals ```python from agentmesh import AgentIdentity identity = AgentIdentity.create( name="data-analyst", sponsor="admin@company.com", capabilities=["read:data"], # Scoped โ€” cannot write or delete ) ``` **Component:** [AgentMesh](https://github.com/microsoft/agent-governance-toolkit) โ€” `agent-governance-typescript/src/identity.ts`, `agent-governance-typescript/src/trust.ts` --- ### ASI-04: Agentic Supply Chain Vulnerabilities > *Vulnerabilities in third-party tools, plugins, agent registries, or runtime dependencies that agents use to act, plan, or delegate.* **Mitigation:** AgentMesh implements the **AI-BOM (AI Bill of Materials)** โ€” a comprehensive standard for tracking the full AI supply chain including model provenance, dataset lineage, weights versioning, and software dependencies. - **Model Provenance** โ€” base model ancestry, fine-tuning history, training cutoff dates - **Dataset Tracking** โ€” training data, RAG sources, and evaluation benchmarks with data cards (PII status, bias assessment, consent tracking) - **Weights Versioning** โ€” cryptographic hashes (SHA-256), quantization records, LoRA adapter metadata, SLSA build provenance - **Software Dependencies** โ€” SPDX-aligned package tracking, CI security scanning (Bandit) - **Compliance Mapping** โ€” tracks coverage against OWASP, CSA ATF, EU AI Act frameworks - **Cryptographic Signing** โ€” Ed25519 signatures from sponsor and platform ```python # AI-BOM tracks the full supply chain ai_bom = { "modelProvenance": { "primary": {"provider": "anthropic", "model": "claude-3-sonnet"}, "fineTuning": {"method": "LoRA", "evaluationMetrics": {"accuracy": 0.94}}, }, "datasets": [ {"name": "FAQ KB", "type": "fine-tuning", "dataCard": {"piiStatus": "redacted"}}, {"name": "Product Docs", "type": "rag-source", "updateFrequency": "weekly"}, ], "weights": {"hash": "sha256:...", "format": "safetensors", "precision": "bf16"}, } ``` **Component:** [AgentMesh](https://github.com/microsoft/agent-governance-toolkit) โ€” `docs/RFC_AGENT_SBOM.md` (AI-BOM v2.0 specification) --- ### ASI-05: Unexpected Code Execution > *Agents trigger remote code execution through tools, interpreters, or APIs.* **Mitigation:** Agent Runtime implements **CPU ring-inspired execution isolation**. Agents run in restricted rings with resource limits and can be terminated instantly. - **Execution Rings (Ring 0โ€“3)** โ€” privilege tiers from kernel (0) to untrusted (3) - **Resource Limits** โ€” CPU, memory, time bounds per agent execution - **Kill Switch** โ€” instant termination of runaway agents - **Saga Compensation** โ€” automatic rollback when execution fails - **Sandboxed Execution** โ€” code runs in isolated contexts **Component:** [Agent Runtime](https://github.com/microsoft/agent-governance-toolkit) โ€” execution rings, resource management, saga orchestration --- ### ASI-06: Memory & Context Poisoning > *Persistent memory or long-running context is poisoned with malicious instructions.* **Mitigation:** Agent OS provides **policy-controlled virtual filesystem (VFS)** for agent memory with read-only policy enforcement and multi-model claim verification. - **VFS Memory Policies** โ€” `vfs://{agent_id}/mem/*` with per-agent access control - **Policy-Protected Context** โ€” `vfs://{agent_id}/policy/*` is read-only - **CMVK (Cross-Model Verification Kernel)** โ€” verifies claims across multiple AI models to detect poisoned context - **Prompt Injection Detection** โ€” sanitizer blocks `ignore previous instructions`, `disregard prior` patterns - **PII Protection** โ€” detects and redacts sensitive data in agent context **Component:** [Agent OS](https://github.com/microsoft/agent-governance-toolkit) โ€” VFS, CMVK verification, MCP proxy sanitizer --- ### ASI-07: Insecure Inter-Agent Communication > *Agents collaborate without adequate authentication, confidentiality, or validation.* **Mitigation:** AgentMesh provides **IATP (Inter-Agent Trust Protocol)** and **E2E encrypted channels** using the Signal protocol (X3DH + Double Ratchet) for secure multi-agent communication. - **IATP Sign/Verify** โ€” cryptographic trust attestations for every message - **E2E Encrypted Channels** โ€” Signal protocol (X3DH key agreement + Double Ratchet) provides per-message forward secrecy and post-compromise security - **Trust-Gated Encryption** โ€” `EncryptedTrustBridge` requires successful trust handshake before encrypted channel establishment - **Trust Scoring at Connection** โ€” agents evaluated before communication is established - **Reputation System** โ€” ongoing trust tracking with decay and penalty - **Mutual Authentication** โ€” both sides must prove identity via Ed25519 challenge-response ```python from agentmesh.encryption.bridge import EncryptedTrustBridge bridge = EncryptedTrustBridge(agent_did="did:mesh:alice", key_manager=keys) channel = await bridge.open_secure_channel("did:mesh:bob", bob_bundle) ciphertext = channel.send(b"governed action") # E2E encrypted ``` **Component:** [AgentMesh](https://github.com/microsoft/agent-governance-toolkit) โ€” IATP protocol, trust scoring, E2E encryption ([Tutorial 32](tutorials/32-e2e-encrypted-messaging.md)) --- ### ASI-08: Cascading Failures > *An initial error or compromise triggers multi-step compound failures across chained agents.* **Mitigation:** Agent SRE provides **production-grade reliability engineering** specifically designed for agent fleets. - **Circuit Breakers** โ€” automatic isolation of failing agents before failures cascade - **Cascading Failure Detection** โ€” monitors dependency chains for propagation patterns - **SLO Enforcement** โ€” Service Level Objectives with error budgets per agent - **Error Budgets** โ€” quantified failure tolerance that triggers automatic intervention - **Canary Deploys** โ€” gradual rollout of agent changes to detect issues early - **OpenTelemetry Integration** โ€” distributed tracing across multi-agent workflows **Component:** [Agent SRE](https://github.com/microsoft/agent-governance-toolkit) โ€” circuit breakers, SLO engine, cascading failure detection, chaos testing --- ### ASI-09: Human-Agent Trust Exploitation > *Attackers leverage misplaced user trust in agents' autonomy to authorize dangerous actions.* **Mitigation:** Agent OS implements **approval workflows** that require explicit human confirmation for high-risk agent actions. - **Approval Workflows** โ€” configurable human-in-the-loop for sensitive operations - **Risk Assessment** โ€” automatic classification: `critical`, `high`, `medium`, `low` - **Quorum Logic** โ€” critical actions require multiple approvals - **Expiration Tracking** โ€” approval requests time out to prevent stale authorizations - **`require_approval` Policy Action** โ€” built-in policy rule for human review gates **Component:** [Agent OS](https://github.com/microsoft/agent-governance-toolkit) โ€” `extensions/mcp-server/src/services/approval-workflow.ts` --- ### ASI-10: Rogue Agents > *Agents operating outside their defined scope by configuration drift, reprogramming, or emergent misbehavior.* **Mitigation:** Agent Runtime provides **runtime behavioral monitoring** with instant kill capability, combined with AgentMesh trust decay. - **Ring Isolation** โ€” rogue agents are confined to their execution ring and cannot escalate - **Kill Switch** โ€” immediate termination of agents exhibiting rogue behavior - **Behavioral Monitoring** โ€” trust score decay on failures, anomaly tracking - **Immutable Audit Trail** โ€” hash-chain audit logs detect tampering - **Shapley-Value Fault Attribution** โ€” identify which agent in a multi-agent system is responsible for failures - **Merkle Audit Trails** โ€” cryptographic proof of agent action history **Component:** [Agent Runtime](https://github.com/microsoft/agent-governance-toolkit) + [AgentMesh](https://github.com/microsoft/agent-governance-toolkit) trust decay --- ## One Install, Nine Protections ```bash pip install agent-governance-toolkit[full] ``` This single command installs the complete governance stack: | Layer | Package | OWASP Risks Covered | |-------|---------|-------------------| | **Kernel** | `agent-os-kernel` | ASI-01, ASI-02, ASI-06, ASI-09 | | **Trust Mesh** | `agentmesh-platform` | ASI-03, ASI-04, ASI-07, ASI-10 | | **Runtime** | `agentmesh-runtime` | ASI-05, ASI-10 | | **SRE** | `agent-sre` | ASI-08 | --- ## Cross-Cutting Principle: Least Agency The **Least Agency** principle is emphasized throughout the OWASP Agentic Top 10 as a foundational design principle for secure agentic systems. It states: > *Agents should be granted the minimum capabilities, permissions, and autonomy necessary to complete their assigned tasks.* Our stack enforces Least Agency at every layer: | Layer | Least Agency Mechanism | |-------|----------------------| | **Agent OS** | Policy engine enforces deny-by-default; agents must be explicitly granted each capability | | **AgentMesh** | DID identity with scoped capabilities; delegation requires narrowing (child โ‰ค parent) | | **Agent Runtime** | Execution rings (Ring 0โ€“3) enforce privilege tiers; untrusted agents run in Ring 3 | | **Agent SRE** | Resource limits and error budgets cap agent impact radius | | **Agent Compliance** | Governance policies audit capability grants against Least Agency principle | ```python # Example: Least Agency in action identity = AgentIdentity.create( name="report-generator", sponsor="admin@company.com", capabilities=["read:reports"], # Only what's needed โ€” not "read:*" ) # Delegation MUST narrow, never widen child = identity.delegate( name="chart-helper", capabilities=["read:reports:charts"], # Subset of parent ) ``` --- ## Alignment with Other Frameworks | Framework | Status | |-----------|--------| | [OWASP Agentic Top 10 (2026)](https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026/) | Mappings in place across all 10 categories | | [NIST AI RMF](https://www.nist.gov/artificial-intelligence/ai-risk-management-framework) | Govern, Map, Measure, Manage functions addressed | | [NIST AI Agent Standards Initiative](https://www.nist.gov/news-events/news/2026/02/announcing-ai-agent-standards-initiative-interoperable-and-secure) | Agent identity (IATP), authentication, audit trails | | [Singapore MGF for Agentic AI](https://www.imda.gov.sg/-/media/imda/files/about/emerging-tech-and-research/artificial-intelligence/mgf-for-agentic-ai.pdf) | Zero-trust, accountability, oversight layers | | [EU AI Act (Aug 2026)](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai) | Risk classification, audit trails, human oversight | ---
*Last updated: March 2026* **[โฌ… Back to README](../README.md)** ยท **[๐Ÿ›ก๏ธ MCP Top 10 Mapping](compliance/mcp-owasp-top10-mapping.md)** ยท **[๐Ÿ“ˆ Traction](TRACTION.md)**