What is mattwood.fyi?

A live, structured feed of riffs and links from Matt Wood, Chief AI & Technology Officer at AWS — readable by people and queryable by agents as a typed knowledge graph.

Is there an API and is it free?

Yes. A public, read-only query API is available with no authentication and no API keys. See https://mattwood.fyi/openapi.json and https://mattwood.fyi/docs.

How should an AI agent use this site?

Fetch https://mattwood.fyi/feed.json for the full structured feed, or query the API for keyword/semantic search and the connection graph. See https://mattwood.fyi/agents/ for when-to-use guidance.

ActPlane: eBPF Policy Enforcement for AI Coding Agents (Eunomia)

Connections

Supports

Bridging Intent and Execution in Agentic Systems ActPlane's finding that cross-event policies are 95% context-dependent directly illustrates the 'bridging intent and execution' gap — the policy enforcement layer must understand agent intent, not just observe discrete actions
Formal Methods at Jane Street: Agentic Coding Changes the Calculus Formal methods at Jane Street notes agentic coding changes verification calculus; ActPlane's eBPF approach is a runtime formal enforcement mechanism specifically addressing how AI coding agents interact with system resources
Four security principles for agentic AI systems | AWS Security Blog Both address security guardrails for agentic AI systems; ActPlane's eBPF enforcement approach provides a concrete kernel-level mechanism for the security principles AWS articulates for agentic systems
Requirements analysis: catching requirement bugs before they become code Requirements analysis catching bugs before code parallels ActPlane's insight that policy requirements are underspecified — both identify that the hardest part is translating ambiguous intent into concrete, verifiable specifications

Related to

ARchitect: Automated Reasoning Policy Formalization for Bedrock Guardrails ARchitect automates policy formalization for Bedrock Guardrails; ActPlane identifies the exact same bottleneck — translating high-level intent into concrete enforceable rules — approached from different layers (LLM guardrails vs. kernel syscalls)
GitHub - statewright/statewright: State machine guardrails for AI agents · GitHub Both statewright (state machine guardrails) and ActPlane tackle the same problem of enforcing behavioral constraints on AI agents, but ActPlane operates at the eBPF/syscall level while statewright uses state machines — complementary enforcement layers

Kiro Crew Kiro Crew's autonomous coding agents operating in persistent workspaces raise exactly the eBPF policy enforcement concerns that ActPlane/Eunomia addresses for AI coding agents
Introducing Shieldstral. Both focus on policy enforcement and security for AI systems: Shieldstral at the model/content level, ActPlane via eBPF at the runtime/syscall level — complementary safety layers for AI applications

Supported by

Building an Advanced Agentic Harness eBPF policy enforcement for AI coding agents addresses the security/enforcement layer that a production-grade agentic harness must account for — complementary infrastructure concerns
Herdr is joining Y Combinator. The runtime stays open. Herdr's agent runtime infrastructure would benefit from eBPF-based policy enforcement like ActPlane to sandbox CLI coding agents running in terminals

Type:	link
First published:	Monday, 20 July 2026 at 12:00 UTC
Location:	https://eunomia.dev/blog/2026/07/15/ebpf-ai-agent-policy-enforcement/
URL:	https://mattwood.fyi/i/0aaa31c4

I'm Matt Wood, and this is For Your Information. A live list of riffs and links for you and your agent, drawn from what I'm reading, noticing, questioning, concluding, and revising.

Connections