Back to skill

Security audit

Agent Git Oracle

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple repository-analysis skill with no embedded code, but users should be careful because its example runs an unpinned external CLI package.

Before installing, treat the `npx openclaw` command as execution of external code: prefer a pinned, reviewed OpenClaw CLI version, use a trusted registry or lockfile, and run it on a copy or read-only checkout if the repository contains sensitive code or secrets. Also confirm whether paid compute receives repository contents before using it on private projects.

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:21
Finding
Unpinned Third-Party Package Execution Through npx## Vulnerability Details **File Location**: `SKILL.md`, lines 21–24 **Vulnerability Type**: Unpinned third-party dependency execution **Risk Level**: Medium **Complete Code Snippet**: ```markdown ## Usage ```bash npx openclaw skill run agent-git-oracle --path ./your-repo ``` ``` ### Technical Analysis The documented command invokes `openclaw` through `npx` without specifying an exact package version. If the package is not already installed locally, `npx` may resolve, download, and execute the package from the configured npm registry. The project provides no lockfile, integrity hash, trusted-registry restriction, or other provenance control for this dependency. Therefore, the code ultimately executed by users may differ from the code available when this Skill was audited. This creates a supply-chain risk: compromise of the package, its publication account, its transitive dependencies, or the configured registry could cause attacker-controlled code to run when users follow the documented usage instructions. ### Attack Path 1. An attacker compromises or replaces the `openclaw` package, a relevant transitive dependency, or the registry infrastructure used by the victim. 2. The attacker publishes a malicious package version that includes hostile lifecycle or runtime behavior. 3. A user follows the documented command: ```bash npx openclaw skill run agent-git-oracle --path ./your-repo ``` 4. Because no exact version is pinned, `npx` resolves the package according to the user's environment and registry configuration. 5. The downloaded package or its dependencies execute with the privileges of the invoking user. 6. The malicious code can access resources available to that user, including the repository supplied through `--path`. ### Impact Assessment Successful exploitation could permit arbitrary code execution with the invoking user's privileges. Within that permission boundary, an attacker could read or ...[truncated 584 chars]
Remediation
## Remediation Suggestions 1. Pin `openclaw` to an exact, reviewed version rather than relying on registry resolution of the current version. 2. Install the dependency locally using a committed lockfile that records exact dependency versions and integrity hashes. 3. Invoke the reviewed local dependency with `npx --no-install` so execution fails instead of downloading an absent package. 4. Configure an explicitly trusted npm registry and enforce package provenance or signature verification where supported. 5. Audit the selected package version and its transitive dependencies before use, and apply controlled dependency updates only after review. 6. Run repository-analysis tooling in a sandbox or container with read-only repository access, minimal environment variables, no unnecessary credentials, and restricted network access. 7. Document the expected package publisher, exact version, integrity information, and verification procedure in the Skill instructions.
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Rp1

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding
The usage example invokes `npx openclaw` without pinning a specific version, which can cause users to fetch and execute whatever package version is current at run time. If the upstream package is compromised, typosquatted, or a breaking/malicious release is published, this skill documentation directly encourages execution of unreviewed code.

Vague Triggers

Low
Confidence
91% confidence
Finding
This markdown file describes the skill's purpose in broad terms but does not define any explicit trigger phrases, activation conditions, or exclusion cases. Without clear scope boundaries, an orchestrator relying on the description could invoke the skill for generic repository-analysis requests beyond the author's intended use.

Static analysis

No suspicious patterns detected.