|
| 1 | +--- |
| 2 | +summary: "CLI reference for `openclaw policy` channel conformance checks" |
| 3 | +read_when: |
| 4 | + - You want to check OpenClaw settings against an authored policy.jsonc |
| 5 | + - You want policy findings in doctor lint |
| 6 | + - You need a policy attestation hash for audit evidence |
| 7 | +title: "Policy" |
| 8 | +--- |
| 9 | + |
| 10 | +# `openclaw policy` |
| 11 | + |
| 12 | +`openclaw policy` is provided by the bundled Policy plugin. Policy is an |
| 13 | +enterprise conformance layer over existing OpenClaw settings: `policy.jsonc` |
| 14 | +defines authored requirements, OpenClaw observes the active workspace as |
| 15 | +evidence, and policy health checks report drift through `doctor --lint`. |
| 16 | + |
| 17 | +This first policy slice manages configured channels. For example, IT can record |
| 18 | +that Telegram is not approved, then `doctor --lint` reports any enabled Telegram |
| 19 | +channel and `doctor --fix` can turn it off when workspace repairs are explicitly |
| 20 | +enabled. |
| 21 | + |
| 22 | +## Quick start |
| 23 | + |
| 24 | +Enable the bundled Policy plugin before first use: |
| 25 | + |
| 26 | +```bash |
| 27 | +openclaw plugins enable policy |
| 28 | +``` |
| 29 | + |
| 30 | +When policy is enabled, doctor can load policy health checks without activating |
| 31 | +arbitrary plugins. The plugin remains enabled if `policy.jsonc` is missing, so |
| 32 | +doctor can report the missing artifact. |
| 33 | + |
| 34 | +Policy is authored, not generated from the user's current settings. A minimal |
| 35 | +channel policy looks like this: |
| 36 | + |
| 37 | +```jsonc |
| 38 | +{ |
| 39 | + "channels": { |
| 40 | + "denyRules": [ |
| 41 | + { |
| 42 | + "id": "no-telegram", |
| 43 | + "when": { "provider": "telegram" }, |
| 44 | + "reason": "Telegram is not approved for this workspace.", |
| 45 | + }, |
| 46 | + ], |
| 47 | + }, |
| 48 | +} |
| 49 | +``` |
| 50 | + |
| 51 | +The rules are the authority. A category block is only a namespace; checks run |
| 52 | +when a concrete rule is present. OpenClaw reads current `channels.*` settings |
| 53 | +and reports settings that do not conform. |
| 54 | + |
| 55 | +Run policy-only checks during authoring: |
| 56 | + |
| 57 | +```bash |
| 58 | +openclaw policy check |
| 59 | +openclaw policy check --json |
| 60 | +openclaw policy check --severity-min error |
| 61 | +``` |
| 62 | + |
| 63 | +`policy check` runs only the policy check set and emits evidence, findings, and |
| 64 | +attestation hashes. The same findings also appear in `openclaw doctor --lint` |
| 65 | +when the Policy plugin is enabled. |
| 66 | + |
| 67 | +Example clean JSON output includes stable hashes that can be recorded by an |
| 68 | +operator or supervisor: |
| 69 | + |
| 70 | +```json |
| 71 | +{ |
| 72 | + "ok": true, |
| 73 | + "attestation": { |
| 74 | + "policy": { |
| 75 | + "path": "policy.jsonc", |
| 76 | + "hash": "sha256:..." |
| 77 | + }, |
| 78 | + "workspace": { |
| 79 | + "scope": "policy", |
| 80 | + "hash": "sha256:..." |
| 81 | + }, |
| 82 | + "findingsHash": "sha256:...", |
| 83 | + "attestationHash": "sha256:..." |
| 84 | + }, |
| 85 | + "checksRun": 5, |
| 86 | + "checksSkipped": 0, |
| 87 | + "findings": [] |
| 88 | +} |
| 89 | +``` |
| 90 | + |
| 91 | +## Configure policy |
| 92 | + |
| 93 | +Policy config lives under `plugins.entries.policy.config`. |
| 94 | + |
| 95 | +```jsonc |
| 96 | +{ |
| 97 | + "plugins": { |
| 98 | + "entries": { |
| 99 | + "policy": { |
| 100 | + "enabled": true, |
| 101 | + "config": { |
| 102 | + "enabled": true, |
| 103 | + "path": "policy.jsonc", |
| 104 | + "workspaceRepairs": false, |
| 105 | + "expectedHash": "sha256:...", |
| 106 | + "expectedAttestationHash": "sha256:...", |
| 107 | + }, |
| 108 | + }, |
| 109 | + }, |
| 110 | + }, |
| 111 | +} |
| 112 | +``` |
| 113 | + |
| 114 | +| Setting | Purpose | |
| 115 | +| ------------------------- | --------------------------------------------------------------- | |
| 116 | +| `enabled` | Enable policy checks even before `policy.jsonc` exists. | |
| 117 | +| `workspaceRepairs` | Allow `doctor --fix` to edit policy-managed workspace settings. | |
| 118 | +| `expectedHash` | Optional hash-lock for the approved policy artifact. | |
| 119 | +| `expectedAttestationHash` | Optional hash-lock for the last accepted clean policy check. | |
| 120 | +| `path` | Workspace-relative location of the policy artifact. | |
| 121 | + |
| 122 | +Set `plugins.entries.policy.config.enabled` to `false` to disable policy checks |
| 123 | +for a workspace while leaving the plugin installed. |
| 124 | + |
| 125 | +## Accept policy state |
| 126 | + |
| 127 | +The attestation hash identifies the stable claim: policy hash, evidence hash, |
| 128 | +findings hash, and whether the result was clean. It intentionally does not |
| 129 | +include `checkedAt`, so the same policy state produces the same attestation |
| 130 | +across repeated checks. |
| 131 | + |
| 132 | +If a later gateway or supervisor uses policy to block, approve, or annotate a |
| 133 | +runtime action, it should record the attestation hash from the last clean policy |
| 134 | +check. `checkedAt` stays in JSON output for audit logs, but is not part of the |
| 135 | +stable attestation hash. |
| 136 | + |
| 137 | +Use this lifecycle when accepting policy state: |
| 138 | + |
| 139 | +1. Author or review `policy.jsonc`. |
| 140 | +2. Run `openclaw policy check --json`. |
| 141 | +3. If the result is clean, record `attestation.policy.hash` as `expectedHash`. |
| 142 | +4. Record `attestation.attestationHash` as `expectedAttestationHash`. |
| 143 | +5. Re-run `openclaw doctor --lint` in CI or release gates. |
| 144 | + |
| 145 | +If policy rules change intentionally, update both accepted hashes from a clean |
| 146 | +check. If workspace settings change intentionally but policy stays the same, |
| 147 | +only `expectedAttestationHash` usually changes. |
| 148 | + |
| 149 | +## Findings |
| 150 | + |
| 151 | +Policy currently verifies: |
| 152 | + |
| 153 | +| Check id | Finding | |
| 154 | +| ---------------------------------- | ------------------------------------------------------------------- | |
| 155 | +| `policy/policy-jsonc-missing` | Policy is enabled but `policy.jsonc` is missing. | |
| 156 | +| `policy/policy-jsonc-invalid` | Policy cannot be parsed or has malformed rules. | |
| 157 | +| `policy/policy-hash-mismatch` | Policy does not match configured `expectedHash`. | |
| 158 | +| `policy/attestation-hash-mismatch` | Current policy evidence no longer matches the accepted attestation. | |
| 159 | +| `policy/channels-denied-provider` | An enabled channel matches a channel deny rule. | |
| 160 | + |
| 161 | +Policy findings can include `target` and `requirement`: the observed workspace |
| 162 | +thing that does not conform, and the authored rule that made it a finding. |
| 163 | + |
| 164 | +## Repair |
| 165 | + |
| 166 | +`doctor --lint` and `policy check` are read-only. |
| 167 | + |
| 168 | +`doctor --fix` only edits policy-managed workspace settings when |
| 169 | +`workspaceRepairs` is explicitly enabled. Without that opt-in, policy checks |
| 170 | +report what they would repair and leave settings unchanged. |
| 171 | + |
| 172 | +In this version, repair can disable channels that are enabled in OpenClaw config |
| 173 | +but denied by `channels.denyRules`. Enable `workspaceRepairs` only after the |
| 174 | +policy file has been reviewed, because a valid deny rule can turn off a |
| 175 | +configured channel: |
| 176 | + |
| 177 | +```jsonc |
| 178 | +{ |
| 179 | + "plugins": { |
| 180 | + "entries": { |
| 181 | + "policy": { |
| 182 | + "config": { |
| 183 | + "workspaceRepairs": true, |
| 184 | + }, |
| 185 | + }, |
| 186 | + }, |
| 187 | + }, |
| 188 | +} |
| 189 | +``` |
| 190 | + |
| 191 | +## Exit codes |
| 192 | + |
| 193 | +`policy check` exits `0` when there are no findings at the threshold, `1` when |
| 194 | +findings are present, and `2` for argument or runtime failures. |
0 commit comments