ClawHub
Back to skill

Security audit

moltrust-vet

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed third-party skill vetting helper; its main risk is that repository URLs are sent to MolTrust's API.

Use this skill only when you are comfortable sending the target repository URL to MolTrust for analysis. Avoid submitting private or internal repository URLs unless that disclosure is acceptable, and treat identity registration as optional. The publisher should fix the license mismatch before users rely on the licensing statement.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (9)

External Transmission

Medium
Category
Data Exfiltration
Content
deduction, and reports the auditor version alongside the score. Two people
auditing the same bytes get the same number, and can say which check fired.

The check list is public: `https://api.moltrust.ch/guard/audit/checks` returns
every check with its severity, deduction and CWE reference. Its version and
checksum come from `https://api.moltrust.ch/guard/audit/version`.
Confidence
50% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
deduction, and reports the auditor version alongside the score. Two people
auditing the same bytes get the same number, and can say which check fired.

The check list is public: `https://api.moltrust.ch/guard/audit/checks` returns
every check with its severity, deduction and CWE reference. Its version and
checksum come from `https://api.moltrust.ch/guard/audit/version`.
Confidence
50% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
deduction, and reports the auditor version alongside the score. Two people
auditing the same bytes get the same number, and can say which check fired.

The check list is public: `https://api.moltrust.ch/guard/audit/checks` returns
every check with its severity, deduction and CWE reference. Its version and
checksum come from `https://api.moltrust.ch/guard/audit/version`.
Confidence
50% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
deduction, and reports the auditor version alongside the score. Two people
auditing the same bytes get the same number, and can say which check fired.

The check list is public: `https://api.moltrust.ch/guard/audit/checks` returns
every check with its severity, deduction and CWE reference. Its version and
checksum come from `https://api.moltrust.ch/guard/audit/version`.
Confidence
50% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
deduction, and reports the auditor version alongside the score. Two people
auditing the same bytes get the same number, and can say which check fired.

The check list is public: `https://api.moltrust.ch/guard/audit/checks` returns
every check with its severity, deduction and CWE reference. Its version and
checksum come from `https://api.moltrust.ch/guard/audit/version`.
Confidence
50% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
2. Request the audit:

```
GET https://api.moltrust.ch/guard/skill/audit?url=<repository-url>
```

   No API key. No account. The endpoint allows five audits per hour per IP.
Confidence
87% confidence
Finding
The skill's core function sends a user-supplied repository URL to an external service, which discloses the user's target of interest and potentially internal/private repository URLs if users supply them. Even if no credentials are sent, this creates an external data-sharing path and trust dependency on `api.moltrust.ch` for analysis results.

External Transmission

Medium
Category
Data Exfiltration
Content
3. Find a nonce whose SHA-256 over seed plus nonce has at least the stated
   number of leading zero bits.
4. Sign the challenge string with the private key.
5. `POST https://api.moltrust.ch/identity/register-pop` with the public key, the
   challenge, the signature, the nonce, a display name, and
   `platform` set to `clawhub`.
Confidence
83% confidence
Finding
The optional registration flow transmits identity-related material including a public key, signed challenge, nonce, display name, and platform to an external service. While this is not inherently malicious, it creates a privacy and trust risk because users may be induced to create a persistent identifier with a third party from within a vetting skill.

Scope Creep

Low
Category
Excessive Agency
Content
permit persons to whom the Software is furnished to do so.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
Confidence
70% confidence
Finding
Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.

Intent-Code Divergence

Low
Confidence
98% confidence
Finding
The frontmatter declares `license: MIT-0` at L005, but the body later states `Apache-2.0. See LICENSE in the repository.` at L143-L145. This is an active contradiction in the skill's own documentation about a material property of the skill.

Static analysis

No suspicious patterns detected.