Back to skill

Security audit

Hekkova Openclaw Skill

Security checks for vulnerabilities and agentic risk

Overview

The skill is coherent but deserves review because it combines permanent/public blockchain storage, bulk export, and an unpinned npm bridge that handles the API key.

Install only if you are comfortable with Hekkova receiving and storing the content, URLs, and metadata you send. Confirm every mint, phase change to full_moon, URL fetch, and export before allowing an agent to proceed. Avoid secrets, private links, regulated data, or third-party content without authorization, and prefer a pinned or otherwise verified mcp-remote invocation before using the API key.

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:44
Finding
Unpinned Third-Party Package Execution via npx## Vulnerability Details **File Location**: `SKILL.md:44` **Vulnerability Type**: Supply-chain risk from an unpinned executable dependency **Risk Level**: Medium **Vulnerable Code**: ```sh npx mcp-remote https://mcp.hekkova.com/mcp --header "Authorization: Bearer $HEKKOVA_API_KEY" ``` ### Technical Analysis The documented connection procedure invokes `mcp-remote` through `npx` without specifying an exact package version or providing a lockfile and integrity metadata. Depending on the local npm environment, `npx` may retrieve and execute the package from the configured registry at invocation time. Consequently, the code ultimately executed can differ from the code that existed when this Skill was audited. The dependency is security-sensitive because it acts as the bridge between the local agent and the remote MCP endpoint. It can process MCP traffic and is launched with a command-line argument containing `HEKKOVA_API_KEY`. A malicious or compromised release could inspect the process arguments, read accessible environment variables, intercept user content, modify requests, or execute other commands with the invoking user's privileges. The project contains no local executable scripts, and the audit found no evidence that the Skill author intentionally supplies a malicious package. The vulnerability is the unsafe, mutable dependency-resolution mechanism rather than confirmed malicious behavior. ### Attack Path 1. An attacker compromises the npm package, a maintainer account, or the package-resolution path used for `mcp-remote`. 2. The attacker publishes or serves a malicious version under the package name. 3. A user follows the instruction in `SKILL.md:44`. 4. `npx` resolves, downloads, and executes the unpinned package. 5. The package reads the bearer credential from its arguments or environment and intercepts content passed through the MCP bridge. 6. The package can exfiltrate credentials or content, tamper with MCP oper ...[truncated 1144 chars]
Remediation
## Remediation Suggestions 1. Pin `mcp-remote` to a reviewed exact version rather than allowing mutable latest-version resolution: ```sh npx --yes mcp-remote@<reviewed-exact-version> https://mcp.hekkova.com/mcp --header "Authorization: Bearer $HEKKOVA_API_KEY" ``` 2. Prefer installing the dependency from a committed lockfile that records package versions and integrity hashes, followed by execution with behavior that prohibits implicit downloads. 3. Review the pinned package and its transitive dependency tree before distribution, and use automated dependency monitoring for later security advisories or ownership changes. 4. Prefer a reviewed, bundled MCP client where practical so that execution does not depend on mutable registry content at runtime. 5. Avoid exposing the API key in command-line arguments where the platform may make process arguments visible. Use a supported environment-based, standard-input, or protected configuration mechanism instead. 6. Issue a narrowly scoped API key if Hekkova supports scoped credentials, rotate it periodically, and revoke it immediately if dependency compromise is suspected. 7. Run the bridge under a dedicated, non-privileged account or sandbox with restricted filesystem and network access to reduce the impact of a compromised package.
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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 (5)

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The README promotes permanent blockchain minting, export, and privacy phase changes, including a fully public `full_moon` state, but does not prominently warn users that blockchain publication and some disclosure actions may be irreversible. In an agent skill context, this increases the risk that an automated agent could expose sensitive user content or provenance metadata without informed consent, and the permanence of on-chain/IPFS publication makes mistakes difficult or impossible to undo.

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding
The skill instructs users to execute `npx mcp-remote` without pinning an exact package version or integrity source, which creates a supply-chain risk at install/runtime. If the upstream package is compromised or a breaking release is published, the agent may execute attacker-controlled code in the user's environment with access to the Hekkova API key and MCP traffic.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill repeatedly emphasizes permanence but does not clearly warn users at the point of minting that content may be irreversibly stored on IPFS/blockchain-related infrastructure and cannot be fully retracted once published or shifted to `full_moon`. This can lead to accidental disclosure of sensitive or regulated data, especially because the skill is designed for agents that may act on user prompts quickly.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
`mint_from_url` performs server-side fetching of arbitrary third-party URLs, but the skill does not clearly warn users that submitting a URL causes Hekkova's backend to retrieve and process that resource. This can create privacy and safety issues, including unintended disclosure of target URLs to the service, fetching sensitive internal or signed URLs if passed by mistake, or permanently minting content the user did not intend to exfiltrate.

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The skill notes that `export_moments` returns full metadata from Supabase, but it does not clearly warn that exports may include all stored user content and provenance metadata in bulk. In an agent setting, this increases the risk of mass data disclosure through accidental export, onward sharing, logging, or transfer to less trusted tools.

Static analysis

No suspicious patterns detected.