Signet Guardian

ReviewAudited by ClawScan on May 10, 2026.

Overview

This payment-guard skill is mostly coherent, but its bundled fallback policy has payments enabled, so it could allow payment-capable skills to proceed before the user has explicitly configured their own policy.

Before installing, set an explicit Signet policy in OpenClaw config and consider changing paymentsEnabled to false until you opt in. Verify that any payment-capable skill you use actually calls and obeys Signet preflight/record results, and remember that the ledger stores local payment history.

Findings (5)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

Payment-capable skills that honor Signet may be told that small payments are allowed using the packaged defaults, not a policy the user explicitly created.

Why it was flagged

The SKILL.md and code describe references/policy.json as the fallback policy after OpenClaw config. Because the bundled fallback enables payments, a fresh or unconfigured setup can return ALLOW under these limits instead of requiring the user to opt in first.

Skill content
"paymentsEnabled": true,
  "maxPerTransaction": 20,
  "maxPerMonth": 500,
  "currency": "GBP",
  "requireConfirmationAbove": 5
Recommendation

Ship the fallback policy with paymentsEnabled set to false, or require an explicit first-run configuration step before any ALLOW result is possible.

What this means

The guard is only effective when every payment-capable skill is integrated correctly.

Why it was flagged

The guard depends on cooperating payment skills. If another skill skips preflight, ignores DENY, or fails to handle CONFIRM_REQUIRED correctly, Signet will not technically block the payment.

Skill content
It does not intercept payments at runtime by itself; payment-capable skills must route through it by contract
Recommendation

Use it only with payment skills that explicitly call signet-preflight before payment and signet-record after payment, and verify confirmation handling.

What this means

Running the edit command executes whatever editor command is configured in the environment.

Why it was flagged

The CLI can spawn a local editor for policy editing. This matches the documented signet-policy --edit behavior, but it is still local command execution influenced by the user's editor setting.

Skill content
const result = spawnSync(editorCmd, [...editorArgs, POLICY_PATH], {
Recommendation

Run signet-policy --edit only in a trusted shell environment and with a trusted EDITOR configuration.

What this means

Local files may accumulate sensitive financial history that other local processes or future reports can read.

Why it was flagged

The skill persistently stores transaction and denial history locally, including payees and payment purposes.

Skill content
Ledger file: `{baseDir}/references/ledger.jsonl`... Plus: amount, currency, payee, purpose.
Recommendation

Treat the ledger as sensitive data, restrict local file access, and periodically review or archive it according to your privacy needs.

What this means

Using the CLI may require installing or running packages from the Node ecosystem.

Why it was flagged

Although registry metadata lists no install spec or required binaries, the documented workflow relies on Node tooling and npm/pnpm packages.

Skill content
Node.js 18+
- `tsx` (used via `npx`, or install locally)

```bash
pnpm install
# or: npm install
```
Recommendation

Install from a trusted source, review package-lock/pnpm-lock contents, and prefer pinned/local dependencies over ad-hoc npx execution where possible.