PaySpawn — On-Chain Spending Limits for AI Agents

SuspiciousAudited by ClawScan on May 10, 2026.

Overview

This skill is coherent and openly about scoped crypto payments, but it gives agents delegated authority to spend real USDC automatically, so users should review limits and approvals carefully.

Install only if you are comfortable giving an agent capped authority to spend real USDC. Start with tiny limits, strict whitelists, a minimal funded pool, credential secrecy, transaction monitoring, and a clear rule that any unexpected payment should require human review.

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

If an agent is misled or misconfigured, it could spend USDC up to the configured limits through automatic API payment flows.

Why it was flagged

The skill explicitly enables automatic agent-triggered payments, which are high-impact financial actions, without documenting a per-payment human approval requirement.

Skill content
`ps.fetch()` handles HTTP 402 payment flows automatically. Agent calls a paid API, PaySpawn pays within the credential limits
Recommendation

Use very low daily and per-transaction caps, strict recipient/API whitelists, monitoring, and an explicit approval policy before agents can trigger payments.

What this means

Anyone or any process with access to the credential may be able to spend funds within its limits.

Why it was flagged

The skill requires a delegated payment credential that can authorize spending within configured limits; this is expected for the product but is not declared in the registry metadata.

Skill content
PAYSPAWN_CREDENTIAL=your_credential_from_dashboard
Recommendation

Keep the credential out of shared logs and repositories, scope it narrowly, rotate it if exposed, and ensure the store metadata is updated to declare the credential requirement.

What this means

Users must trust the npm package and its dependencies to handle payment credentials and transactions correctly.

Why it was flagged

The skill relies on an external SDK install, but the reviewed artifact set contains no package code, lockfile, or pinned package version for verification.

Skill content
npm install @payspawn/sdk
Recommendation

Verify the package provenance, pin versions, review the SDK source and audits if available, and install only from the expected package registry.

NoteHigh Confidence
ASI08: Cascading Failures
What this means

A bad pool setting, compromised workflow, or mistaken automation could affect multiple agents and spend from the shared budget.

Why it was flagged

Fleet provisioning can multiply delegated spending authority across many agents from a shared budget pool.

Skill content
Provision 10 agent credentials in one call
Recommendation

Provision incrementally, keep shared pool budgets small, revoke unused credentials, and monitor spending per agent.

What this means

Users may over-trust the stated protections without independently verifying how the credential and contract limits work.

Why it was flagged

The skill makes strong safety claims about smart-contract-enforced limits, but the supplied artifacts do not include contract code, audits, or SDK implementation details to verify them.

Skill content
Credential limits are enforced by the smart contract, not software
Recommendation

Review the PaySpawn documentation, contract addresses, audits, and SDK behavior before funding credentials with meaningful amounts.