Complement to clawsec: fail-closed dispatch boundary + signed receipt per OpenClaw tool call

[peycheff-com]

clawsec focuses on SOUL.md policy drift and config hardening — solid foundational layer. One gap it leaves open: there's no signed artifact proving which MCP tool calls were actually dispatched (allowed vs blocked) within an OpenClaw session after the session ends.

We're building HELM AI Kernel (OSS, MIT), a dispatch-layer complement:

• Sits between the OpenClaw/NanoClaw agent and its MCP servers via an OpenAI-compatible proxy
• Enforces default-deny policy per tool call (ALLOW/DENY/ESCALATE)
• Quarantines unknown MCP servers until schema and policy approval
• Emits a signed receipt per dispatch decision (independent of the model's output)
• Bundles receipts into an offline-verifiable EvidencePack

The stack together:

• clawsec: SOUL.md drift detection, config hardening, pre-session guardrails
• HELM Kernel: dispatch enforcement + tamper-evident session audit trail

Would you be open to testing HELM in front of one of the OpenClaw variants clawsec supports, to see if default-deny fires on policy-restricted tools and the receipt trail complements your SOUL.md audit?

Local demo (~5 min):

brew install mindburnlabs/tap/helm-ai-kernel
helm-ai-kernel serve --policy ./release.high_risk.v3.toml
helm-ai-kernel boundary status --json
helm-ai-kernel receipts tail --agent agent.demo.exec --server http://127.0.0.1:7714
helm-ai-kernel verify evidence-pack.tar

Repo: https://github.com/Mindburn-Labs/helm-ai-kernel

One feedback question: Does clawsec currently produce any artifact that survives a session and proves which tool calls were dispatched vs blocked — or is that an acknowledged gap in the current security model?

[davida-ps]

Hello Ivan,

this question is interesting and requires some research. I am almost positive currently clawsec does not support nor acknowledge this as a gap. clawsec, more than the suite of skills it is, places strong emphasis on supply chain and skill signing. and often the security skills are a by-product (albeit very welcome).

I am interested in hearing more, in which clawsec skill (or pipeline) would you expect this gap to be acknowledged or remediated?

Best Regards
David.