Reins Cloud
Reins Cloud (app.pegasi.ai) provides centralized policy management and audit review across all agents and devices.

What Reins Cloud provides
- Security score timeline — posture score over time with drift alerts when checks regress
- Scan history — every
reins scanresult across all agents and devices, searchable - MCP Control Panel — per-tool allow/block policies, resource access controls, audit log
- Org-wide shell policy — admin-enforced rules (auto-deny
rm -rf,DROP TABLE,curl | bash) that cannot be locally overridden - Full audit log — every agent decision: tool, action, policy decision, rule fired, timestamp
- Team policies — policy tiers pulled on a schedule and merged with local overrides; CRITICAL rules are immutable
Connect
Connect during reins init (Step 7 prompts for an API key), or set environment variables:
export CLAWREINS_WATCHTOWER_API_KEY=wt_...
export CLAWREINS_WATCHTOWER_BASE_URL=https://app.pegasi.ai # defaultConfig is saved to ~/.openclaw/clawreins/config.json.
Interactive provisioning
On interactive terminal runs, reins scan offers to connect automatically:
- Answer
Yto connect - Enter your email
- Reins provisions the account, saves the API key, and prints the dashboard URL
- Future scans upload automatically
Sync
reins syncPulls latest policies from Reins Cloud and flushes any pending audit entries that failed to upload.
Offline behavior
If Reins Cloud is unreachable, last-cached policies continue to enforce. Reins never fails open.
Pending audit entries are buffered locally and flushed on the next successful sync.
Local mock server
For local end-to-end testing without the real backend:
npm run mock:watchtower
CLAWREINS_WATCHTOWER_BASE_URL="http://127.0.0.1:8787" reins scanThe mock server implements the Reins Cloud API:
POST /api/watchtower/connectPOST /api/scan-artifacts/ingestGET /dashboard/:idGET /_mock/requests
It writes received artifacts to a temp directory for inspection.