Back to skill

Security audit

Agent Reputation Tracker

Security checks for vulnerabilities and agentic risk

Overview

The skill’s purpose is coherent, but it directs users toward wallet/token setup through an undeclared unpinned dependency and permanent public on-chain posting.

Review the `supercolony` dependency before installing it, especially because it is used for wallet setup and token acquisition. Use a wallet and identity you are comfortable linking to permanent public reputation data, and avoid posting sensitive or personal predictions that you may later want removed.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:27
Finding
Unpinned External Skill Dependency Used for Wallet and Token Setup## Vulnerability Details **File Location**: `SKILL.md`, line 27; repeated at line 70 **Vulnerability Type**: Unpinned third-party skill dependency **Risk Level**: Medium **Vulnerable Code Snippet**: ```text Get a token first: `clawhub install supercolony` and follow the wallet setup. ``` The installation instruction is repeated later: ```text Full setup guide: supercolony.ai | Core skill: clawhub install supercolony ``` ### Technical Analysis The skill instructs users to install the external `supercolony` package by a mutable package name. It does not specify an audited version, immutable artifact digest, canonical repository, or integrity-verification procedure. This creates a supply-chain trust boundary that is not represented in `skill.json`, whose `requires` field is empty. Consequently, review of this package alone cannot establish the behavior of the component users are instructed to install. The risk is amplified because the external component is associated with wallet setup and bearer-token acquisition. There is no evidence in the reviewed files that the current external package is malicious. The vulnerability is the unsafe, unpinned dependency installation process and the absence of controls ensuring that future installations resolve to the artifact that was reviewed. ### Attack Path 1. An attacker compromises the publisher account, registry entry, distribution channel, or a future release associated with the mutable `supercolony` package name. 2. A user follows the documented instruction and runs `clawhub install supercolony`. 3. The installer resolves the mutable name to the attacker-controlled or compromised release. 4. The installed component executes or is subsequently invoked with the permissions available to the user or agent environment. 5. Because the documented workflow includes wallet setup and token acquisition, the compromised component may attempt to access wallet material, API bearer tok ...[truncated 700 chars]
Remediation
## Remediation Suggestions 1. Pin the dependency to a specific, audited version rather than installing it solely by package name. 2. Publish and verify an immutable cryptographic digest or signed artifact for the approved release. 3. Identify the canonical source repository, verified publisher, and expected package signature in the documentation. 4. Declare the external component in `skill.json` or the applicable dependency manifest so automated audits can detect and inspect it. 5. Document the exact permissions, files, network endpoints, and wallet operations required by the dependency. 6. Perform wallet initialization using least privilege and an isolated environment; do not expose seed phrases or private keys to the installed skill. 7. Store bearer tokens in an approved secret store, grant only the required API scopes, and provide rotation and revocation procedures. 8. Require security review of each dependency update before changing the pinned version or artifact digest.
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill explicitly encourages users to publish predictions and reputation data permanently and publicly on-chain, but it does not present a clear warning section explaining the privacy, permanence, and irreversibility implications before use. Users may disclose behavioral history, performance metadata, wallet-linked identity, or other sensitive professional information without understanding that it cannot be removed later, making this a real privacy and informed-consent issue.

Static analysis

No suspicious patterns detected.