AGIRAILS Escrow Payments
Analysis
This payment skill is coherent, but it should be reviewed because it can let an agent use wallet keys to move real USDC and maintain payment-related state.
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.
Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.
mode: 'mainnet' ... // One line - creates tx and locks escrow
const result = await client.basic.pay({
to: '0xProviderAddress',
amount: '25.00'The skill documents a direct mainnet payment path that creates a transaction and locks escrow with real funds; this is purpose-aligned but financially high impact.
[0] node | package: @agirails/sdk | creates binaries: actp
The payment and wallet functionality is supplied by an external npm package/CLI; this is expected for the skill but is a dependency users must trust.
TREASURY_DIR="$WORKSPACE/agents/treasury" ... cp "$SKILL_DIR/openclaw/SOUL-treasury.md" "$TREASURY_DIR/SOUL.md"
The setup script creates a persistent Treasury agent workspace and installs agent instructions; this is disclosed and user-directed, but it changes the ongoing agent environment.
Checks whether tool use, credentials, dependencies, identity, account access, or inter-agent boundaries are broader than the stated purpose.
SDK auto-detects: keystore → `ACTP_KEYSTORE_BASE64` → `ACTP_PRIVATE_KEY` → `PRIVATE_KEY`
The skill can use wallet credentials and even a generic PRIVATE_KEY environment variable, while the registry requirement highlights only ACTP_KEY_PASSWORD.
Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.
mkdir -p "$TREASURY_DIR/memory" touch "$TREASURY_DIR/memory/transactions.jsonl"
The setup creates persistent memory/log files for payment history and spending state, which future agent decisions may rely on.
