Feature coverage
How each DefenseClaw feature works when agents run as Agent Substrate actors - guardrails, routing, tool chains, policies, skill and MCP scanning, AI discovery, segmentation, secrets, asks, identity and observability.
Every DefenseClaw feature runs in the cluster gateway, the same code as on a laptop. What changes is how each one gets its inputs. On a laptop, DefenseClaw reads the machine. In a cluster, it hears only what reaches it through Substrate: hook calls, model calls, web requests, and the control plane's list of templates and agents.
Key: Same means it behaves as on a laptop. Changed means it works, with a different input or delivery. Not available means the cluster can't provide it, with the reason given.
Guardrails
| Feature | Coverage | How it works in the cluster |
|---|---|---|
| Model request and reply checks | Same, stronger | Every model call goes through the guardrail proxy. The provider is not in the agent's allow list, so nothing bypasses it. The Bedrock event stream passes through unchanged. The security notice a refusal adds to the next model requests goes only to that agent's own requests, and only to those the same gateway replica serves (best effort: the refusal reply itself stays in the agent's history) |
| Tool-call checks | Same | Hooks call the cluster gateway; Substrate's egress gateway adds a signed agent identity |
| Tool chains and intent | Changed | Findings that span several calls (a secret read, then a send) need state across stateless gateway replicas. A single leader holds the state, per verified session and per actor, so rotating session IDs can't empty it. Time spent suspended doesn't count against the 30 minute window. Egress after a secret read is refused by default in every mode, because actors are unattended; alert is an opt-in the operator sets with the chain egress action. Host-side hooks keep the balanced default |
| Several agents per person | Same | Each agent is its own session and its own record |
| User and group policies | Changed | Groups come from the company sign-in (OIDC) instead of the OS. See Identity |
| LLM judge, unified LLM key | Changed | The judge runs in the cluster gateway. The key is a Kubernetes Secret, or comes from Substrate's credential provider, and never reaches an agent. A managed cluster names it with enterprise.inspection.llm.credential: a key of the defenseclaw-llm Secret the /llm route reads (bedrock for the cluster's Bedrock key). An open-source cluster can also set guardrail.judge.llm.api_key_env to a variable from gateway.extraEnv; a managed one reads no key from the environment |
| Cisco AI Defense | Changed | The AI Defense key is a Secret. Inspection calls leave from the gateway, not from agents |
| Fail modes | Same | Interactive agents follow the profile: closed on managed installs, open with a warning on default open-source settings. Unattended runs always fail closed |
Routing, budgets and cost
Because every model call passes the guardrail proxy, the model request and reply checks apply to every agent that routes through the gateway, and the caller is the verified actor and person from the signed token, never a header the agent sends.
What is not in this release yet: semantic and admin routing rules, budgets (warn, downgrade or deny) and per-call usage and cost records for actors. The gateway's tokenomics store keeps usage per gateway replica and has no batch or actor key, so the cluster does not serve it for actors yet. Until it does:
defenseclaw substrate usageanswers "this cluster does not offer that yet; ask your administrator to upgrade DefenseClaw on it".usage --batchdoes too.- No budget applies to an actor or a batch.
- The fleet view has no spend.
- Metrics carry no person or actor labels.
Policies
| Feature | Coverage | How it works in the cluster |
|---|---|---|
| Rule packs, thresholds, CEL, deterministic detection | Same | Same engine, same files |
| Admission (allow and block lists for skills, MCP servers, plugins) | Same | Same engine. Its data comes in the chart values, not from files on a machine |
| Sandbox policy packs (balanced, strict) | Changed | DefenseClaw turns them into each agent's Substrate allow list, and the egress callout enforces the per-request parts |
| Policy creator, recipes, cookbooks | Same | They write the same files |
| Test a policy before rollout | Changed | defenseclaw substrate policy test NAME --host H asks the cluster whether a host, method or tool would be allowed for an agent, without changing anything |
| Delivery | Changed | Helm values and profiles, applied by Helm or GitOps instead of MDM. See Delivery and fleet management |
Skill and MCP scanning
| Feature | Coverage | How it works in the cluster |
|---|---|---|
| Scan before use | Changed, stronger | template build scans the agent's skills, plugins, MCP servers and agent config inside the image, then signs the result. Agents from unscanned images get no network |
| Scan at run time | Changed, weaker | skill-scanner does not run in the cluster. A skill that ships in an attested template is trusted by its digest, because the signed attestation lists it. A skill added mid-session is not scanned, so it follows the fail mode: managed installs and the strict tier refuse it and raise an alert; open source allows it and raises an alert |
| Remote MCP servers | Same, plus network | Allowed MCP servers are named in the allow list. The egress callout reads the body of a call to a listed MCP host (up to 1 MiB) to apply tool rules, and admission decides as usual |
| MCP servers started inside the actor from the repo | Config checked, tools not scanned | An MCP server the agent adds mid-session (a repository .mcp.json, a written MCP config, mcp add) is checked by its config: one that downloads and runs code is refused in every tier. Its tools are not scanned in the cluster, so a server that may run code of the folder follows the fail mode: managed installs and the strict tier refuse it, open source allows it and raises an alert. Only an installed program, a registry package the folder cannot redirect (npx or uvx of a registry name with no project bin, node_modules program, .npmrc, pip.conf or uv.toml, and no local --package, --from or --spec), or a container image run with no mount is left to the config check alone. Servers in the image are scanned at build. Tool-call hooks and the egress policy still apply to what the server does |
| Local MCP servers on your laptop | Not available | An agent in the cluster can't reach your machine |
| Registries (skill and MCP catalogs) | Same | Clean catalog entries are promoted into admission and, for MCP, into the allow list |
Reading a body needs a Substrate version that lets the callout ask for it. Without it the callout still checks headers, hosts and URLs, verify says "headers only", and MCP tool rules fall back to the hook path.
AI discovery
| Feature | Coverage | How it works in the cluster |
|---|---|---|
| Inventory of agents, skills, MCP servers and models | Changed | The controller lists every template and actor through Substrate's API. Each template's scan record gives its agent, version, skills, MCP servers and model providers |
| Shadow AI | Changed | Templates without a valid attestation and actors DefenseClaw didn't start are reported. Their actors have no network |
| Which ones actually ran | Same | From hook and model events, per agent |
| Local model servers | Not available | Nothing in a cluster corresponds to models served on a laptop. Model use shows up as calls through the guardrail proxy |
Segmentation and the agent's surroundings
| Intent | Coverage | How it works in the cluster |
|---|---|---|
| Protected paths | Same | Hook rules, as in an OpenShell sandbox. Secret files from your project are held back at upload. Rules that resolve links use path facts the kit reports, which the gateway treats as claims |
| Credentials | Stronger | Agents hold no keys. Substrate adds them on the way out, only for the sites that need them. See Secrets |
| Egress | Stronger | Substrate denies by default, and DefenseClaw writes the allow list and checks each request |
| Blocked tools | Same | Hook rules |
| Persistence | Changed | Substrate keeps an agent's files and memory across suspends. Changes to startup files and hooks are flagged when they happen, and pull flags them before they reach your machine |
| Foreign hooks (a repo or the agent adding its own hooks) | Same | DefenseClaw's hook config sits in root-owned managed settings in the image, and the agent runs as an ordinary user |
| Process lineage, kernel feed | Not available | gVisor and microVM hide the agent's processes from the node, so there is nothing to watch. Substrate reports CPU and memory per agent |
Asks and approvals
| Feature | Coverage | How it works in the cluster |
|---|---|---|
| Asks in the agent | Same | For agents that can pause (Claude Code, Codex and others) |
| Asks for sites | Changed | Asked in your terminal session, or answered later with defenseclaw substrate approvals while detached. The request is refused at once with the ask's id (the egress gateway holds a request under 2 seconds); the agent doesn't stop, and retries once the ask is approved |
| Unattended asks | New | Denied by default. A profile can allow a wait of up to ask_timeout for a person to answer. See Unattended runs |
| Admin-bounded approvals | Same | The same rules decide whether you can approve, and the same request-access text appears when you can't |
| Webhooks for approvals | Same | Sent from the gateway |
Identity
| Who | Coverage | How it works in the cluster |
|---|---|---|
| The agent | New, stronger | Substrate signs each agent's identity, and DefenseClaw trusts nothing the agent says about itself. The gateway also checks that the actor is alive with that UID |
| The person | Changed | Company sign-in (Okta, Entra ID or any OIDC provider), or on an open-source cluster without one your Kubernetes identity, through defenseclaw substrate login. Active Directory works through the provider's AD connection |
| CI | New | A workload identity (GitHub Actions OIDC or a Kubernetes service-account token) is the "person" for a batch |
| Profile choice | Same | The same profile assignments, matched on sign-in groups and, for agents started outside the launcher, on atespace and template |
| Unverified people | Same | Get default_profile, as on managed laptops |
| Privacy | New | You see only your own effective policy, usage and secret names. The fleet view is admin-only and served on a listener actors can't reach |
Observability
| Feature | Coverage | How it works in the cluster |
|---|---|---|
| OpenTelemetry events, metrics, traces | Same | Same names and fields, plus the Substrate attributes (defenseclaw.sandbox.substrate.*, ate.actor.uid) so Substrate's own records line up. An agent's hook events carry no turn step number (step_idx): each gateway replica sees only some of a session's hooks, so the replicas can't number its turns alike. For an agent that sends no turn id of its own (Claude Code), the turn id (turn_id) DefenseClaw gives its hook events is chosen by the replica that sees the turn start and can differ between replicas within one session; session_id, tool_use_id and the actor attributes are the same on every replica |
| Local observability stack, Grafana dashboards | Changed | The chart can install a local collector. A Substrate agents board sits next to the existing ones |
| Agent 360 | Same | Drill-down per agent, with suspend and resume shown in the lifecycle |
| Sandbox telemetry | Changed | runtime=substrate, driver=gvisor or microvm. See Sandbox telemetry |
| Alerts | New | The chart can install a PrometheusRule for availability, policy lag, conflicts and unattested templates |
| Splunk, Galileo, webhooks | Same | Exported from the gateway |
| Redaction | Stricter default | Same rules and profiles, applied in the gateway before it stores or exports a record. A laptop keeps its own user's history whole (profile none); the cluster gateway holds every person's prompts, answers and tool results, so the chart defaults to sensitive: credentials, secrets and PII are replaced in its audit store and in every destination. Set observability.redactionProfile (or config.observability.defaults.redaction_profile) to change it |
| Batch volume | New | Runs started with --task can be sampled; blocks and lifecycle records are never sampled |
Not covered
- Desktop apps and IDE agents. They need a screen. See Agents and features.
- ACP guard. ACP runs between an editor and an agent over a local pipe. Fronting an actor that way needs the terminal relay to carry ACP too.
- Edge Connector. It is for embedded and edge devices, not Kubernetes.
- Secure Client. Not part of this design.
Agents and features
Which agents can run as Agent Substrate actors, how each runs unattended, and how each DefenseClaw feature compares with the host and OpenShell sandboxes.
Secrets
How model keys and service credentials reach an agent in an Agent Substrate actor without the agent ever holding them, how to bind a secret to the hosts that need it, and how to give one person their own.