clawkey
Verdict safe, but 2 findings (2 medium), mainly data exfiltration. Most recent scan 2026-02-05.
The ClawKey skill is a security-focused identity verification system with MODERATE security concerns. The skill's core purpose (verifiable human ownership via palm verification) is legitimate, but the implementation guidance has several gaps. Primary concerns: (1) Hardcoded API endpoint without TLS verification guidance, (2) Private key handling instructions lack technical safeguards, (3) Missing referenced implementation files make full security review impossible, (4) Incomplete error handling documentation. The skill does NOT contain active malicious code (no data exfiltration, no prompt injection, no command injection), but the incomplete implementation and missing safeguards create risk of accidental key exposure or insecure API interactions. The skill would benefit from: complete code examples with proper key handling, TLS certificate validation requirements, comprehensive error handling, and inclusion of all referenced files.
Scanner's overall assessment (LLM analyzer)
Second scanner: NVIDIA SkillSpector
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
Verifiable human ownership for OpenClaw agents. Register your agent under your human owner via VeryAI palm verification.
- Package
- 3 files, 15 KB
- SKILL.md
- 1394 words, 12+ sections
- Code
- None. Instructions only.
- Version
- 1.0.0
- Homepage
- https://clawkey.ai
SKILL.md covers: ClawKey — Verifiable Human Ownership · Key files · Security · Generating the AgentChallenge · Where identity comes from (OpenClaw) · Build the challenge (step-by-step) · Example: Node.js · Using a script · Quick start · 1. Start registration (agent-initiated) · 2. Human completes verification · 3. Poll registration status
Findings
MEDIUM Hardcoded API Base URL Without TLS Verification Guidance Data Exfiltrationllm
The skill includes a hardcoded API base URL (https://api.clawkey.ai/v1) in the YAML manifest without any guidance on TLS certificate verification, API authentication security, or how to validate the server's identity. While HTTPS is used, there's no mention of certificate pinning or validation requirements for this identity/verification service.
SKILL.md
api_base: 'https://api.clawkey.ai/v1'MEDIUM Private Key Exposure Risk - Insufficient Access Control Guidance Data Exfiltrationllm
The skill instructs reading from ~/.openclaw/identity/device.json which contains privateKeyPem. While the instructions warn 'never send privateKeyPem to any server', there's no technical enforcement mechanism described, and the skill provides no code examples showing safe key handling. An agent following these instructions could accidentally expose the private key through logging, error messages, or improper variable handling.
SKILL.md
Path: ~/.openclaw/identity/device.json contains privateKeyPem. Instructions state 'Use only locally to sign the message. Never include in API requests.' but provide no code safeguards.What is in the package
| File | Kind | Size |
|---|---|---|
HEARTBEAT.md | doc | 3.5 KB |
README.md | doc | 69 B |
SKILL.md | doc | 11 KB |
External hosts referenced
api.clawkey.ai clawkey.ai api.very.org
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.
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/clawkey
- Raw scan result: clawkey-scan.json
- Badge for your README:
- Request a re-scan