Skip to content

Agent actions allowlist (safety gateway)

SUPERSEDED — 2026-08-03. The 12-line tier-a claim stub at this path has been promoted into a tier-b recipe connection at connections/agent-actions-allowlist/. The legacy tier-a "RoamCore native" claim is honest-upstream-truth: RoamCore ships no native agent-actions engine today; the contract layer + the recipe + the FIVE §8 MANDATORY automations are the canonical RoamCore contribution; the upstream engines are HA core input_boolean + input_text + input_number + input_select + input_datetime + input_button + script helpers (since 2022.x) + the HA core template: sensor wrapper (since 2022.x) + the HA core logbook integration (since 2022.x) + the upstream script: integration (since 2022.x). The single input_boolean.rc_agent_actions_enabled kill switch is already shipped in homeassistant/packages/roamcore_agent_actions.yaml and is preserved verbatim by the slice; the slice ADDS the remaining 10 rc_agent_actions_* contract tiles + the recipe + the smoke + the FIVE §8 MANDATORY automations + the operator-side policy file format (with the EXAMPLE policy file at connections/agent-actions-allowlist/docs/policy.example.yaml). Replaced by the tier-b recipe connection; see connections/agent-actions-allowlist/ README + connections/agent-actions-allowlist/docs/recipe.md. Wave 3 #65, PR #69.

What it does

A safety layer that defaults to deny and only permits explicitly-allowed agent actions, with a kill switch.

Why it’s useful in a van

  • Lets you use automation/agents without fear of unexpected device control
  • Clear boundary between “read-only” and “can change things”

How to install

  • None

  • Design notes: docs/design/agent-actions-allowlist.md

  • HA package: homeassistant/packages/roamcore_agent_actions.yaml
  • (Add safety philosophy + examples)

How it works

What RoamCore does behind the scenes.