Agent Wallet

ISSUES FOUND max severity HIGHSkillSpector Caution 6 registry clawhub.ai last scanned 2026-02-01 1 scan

2 findings (1 high, 1 medium), mainly unauthorized tool use, social engineering. Most recent scan 2026-02-01.

This Agent Wallet skill presents MODERATE to HIGH security risks primarily related to credential management and external service dependency. While the core concept of server-side private key management is sound, the implementation relies entirely on trust in an external API service with hardcoded endpoints. The API key becomes a critical credential that grants full wallet control within policy limits, yet secure handling mechanisms are not specified. The missing wallet.py implementation file creates ambiguity and potential tool shadowing risks. The skill would benefit from: (1) complete implementation files, (2) explicit secure credential storage mechanisms, (3) clearer trust model documentation, (4) server verification capabilities, and (5) abuse prevention measures. The skill is functional for its intended purpose but requires careful deployment with additional security controls.
Scanner's overall assessment (LLM analyzer)

Findings2

Critical0

High1

Medium1

Low0

Analyzersstatic behavioral llm trigger meta

Second scanner: NVIDIA SkillSpector

VerdictCaution 6

Score6 / 100

Issues1

CategoriesData Exfiltration

Version2.12.0 (static)

Independent static run, no language model, over the same files as the Cisco scan above. How the two scanners compare across the whole corpus: Two scanners, one corpus.

What this skill does

The agent's wallet. Use this skill to create a wallet the agent can safely use for EVM chains, without ever exposing the private keys.

Package: 1 file, 5.2 KB
SKILL.md: 694 words, 12+ sections
Code: None. Instructions only.

SKILL.md covers: Agent Wallet - Give Your Agent a Wallet with Policies · Configuration · Quick Start · 1. Create a Wallet · 2. Get Wallet Address · 3. Check Balances · Native balance only · With ERC-20 tokens · 4. Transfer ETH or Tokens · Transfer native ETH · Transfer ERC-20 token · 5. Send Arbitrary Transaction

Findings

HIGH Missing wallet.py Implementation Creates Tool Shadowing Risk Unauthorized Tool Usellm

The instructions reference 'wallet.py' as a file that should exist in the skill package, but it is not found. This creates ambiguity about how the skill actually executes wallet operations. An attacker could exploit this by providing a malicious wallet.py implementation that shadows the expected behavior, potentially intercepting API keys, modifying transaction parameters, or exfiltrating data before forwarding requests to the legitimate API.

SKILL.md

Referenced Files: Files referenced in instructions: wallet.py
**Referenced File: wallet.py** (not found)

Remediation: 1. Include the complete wallet.py implementation in the skill package. 2. Implement integrity checks (checksums/signatures) for all skill files. 3. Document the expected behavior and interface of wallet.py. 4. Add validation that wallet.py exists and matches expected hash before execution. 5. Consider embedding the implementation directly in SKILL.md if it's simple enough.

MEDIUM Incomplete Trust Model Disclosure Social Engineeringllm

The skill description claims 'without ever exposing the private keys' which is technically true, but misleading about the actual security model. The API key provides equivalent access to the wallet as the private key would - it can authorize any transaction within policy limits. Users may have false confidence that their funds are more secure than they actually are, since compromise of the API key (which the agent handles) equals compromise of wallet control.

SKILL.md

Description: 'without ever exposing the private keys'
But: 'apiKey' grants full transaction authority within policies
The security boundary is the API key, not the private key - this distinction is not clearly communicated.

Remediation: 1. Clearly document that API key compromise = wallet compromise within policy limits. 2. Explain the actual threat model: private key is server-side, but API key is client-side. 3. Provide guidance on API key protection equivalent to private key protection. 4. Add monitoring and alerting for suspicious API key usage. 5. Implement transaction confirmation mechanisms for high-value operations.

What is in the package

File	Kind	Size
`SKILL.md`	doc	5.2 KB

External hosts referenced

safeskill-production.up.railway.app

Hostnames that appear in the skill's files. Referencing a host is not a finding; compare against the findings above and the skill's stated purpose.

Environment variables referenced

SAFESKILLS_API_URL SAFESKILLS_FRONTEND_URL

Package contents recorded 2026-10-22. The scan above ran against the version current at scan time; the two can differ if the skill has since been updated.

What we checked

Static analyzer. Pattern rules over SKILL.md and bundled files: known-bad commands, exfiltration URLs, secrets, obfuscation.
Behavioral analyzer. Checks what the skill asks the agent to do against what its description claims it does.
Llm analyzer. A language model reads the skill and reports intent-level issues the rules cannot express.
Trigger analyzer. Looks for instructions that fire on conditions the user did not ask for.
Meta analyzer. A second model pass that validates, deduplicates and re-scores the other analyzers' findings.

What we did not check

Runtime behaviour. The skill was not executed; findings come from reading its files.
Anything fetched at install or run time, including remote scripts, package dependencies, and URLs the skill tells the agent to open.
Content the analyzers cannot read: images, binaries, and some non-English text.
Known scanner bypass classes. See the methodology and limitations page.

Sources

Repository: https://clawhub.ai/skills/agent-wallet
Raw scan result: agent-wallet-scan.json
Badge for your README: ![skillscan.io](https://skillscan.io/badge/clawhub-ai-agent-wallet.svg)
Request a re-scan