Back to skill

Security audit

context-manager-agentic

Security checks for vulnerabilities and agentic risk

Overview

The skill has a coherent context-management purpose, but installing it depends on mutable external code that was not included for review.

Review the referenced GitHub repository and its npm dependency tree before installing. Prefer a pinned commit or signed release with a lockfile, and run installation in a sandbox with only the files and credentials needed for the context-management task.

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:26
Finding

Unpinned Remote Repository and Unaudited npm Installation

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 26–27
Vulnerability Type: Unpinned remote source and unsafe dependency installation
Risk Level: Medium

bash
git clone https://github.com/Tryboy869/context-manager-agentic
cd context-manager-agentic && npm install

Technical Analysis

The skill delegates its functionality to a mutable external GitHub repository and does not pin the cloned content to a reviewed commit SHA or signed release. Consequently, the code retrieved by users can differ from the code that existed when the skill documentation was reviewed.

The subsequent npm install command resolves packages using manifests supplied by that remote repository. It may also execute npm lifecycle scripts, such as preinstall, install, and postinstall, with the permissions of the installing user. The local skill package does not contain the referenced implementation, dependency manifest, or lockfile, so the dependency graph, lifecycle scripts, and effective executable behavior cannot be audited from this artifact.

This creates a supply-chain trust gap: compromise of the GitHub account, repository, release process, npm dependencies, or transitive dependencies could introduce executable code without any modification to the reviewed SKILL.md file.

Attack Path

  1. An attacker compromises the referenced GitHub repository, its owner account, or one of its npm dependencies.
  2. The attacker adds a malicious dependency, lifecycle script, or modification to cli.js.
  3. A user follows the documented instructions and clones the repository's current mutable state.
  4. The user runs npm install, causing attacker-controlled package code or lifecycle scripts to execute, or later invokes the modified cli.js.
  5. The malicious code runs with the user's privileges and can access resources available to that process.

Impact Assessment

Successful exploitation could permit arbitrary cod ...[truncated 715 chars]

Remediation
View remediation

Remediation Suggestions

  1. Include the complete implementation and dependency metadata in the skill package so that the effective behavior can be reviewed together with SKILL.md.
  2. Pin the external repository to a specific, reviewed commit SHA rather than cloning the default branch.
  3. Prefer a signed release artifact and verify its cryptographic signature or published checksum before use.
  4. Commit a reviewed package-lock.json and replace npm install with npm ci to enforce the locked dependency graph.
  5. Audit all direct and transitive dependencies for known vulnerabilities, suspicious maintainers, dependency confusion, and typosquatting.
  6. Review all npm lifecycle scripts and use npm ci --ignore-scripts where installation scripts are unnecessary.
  7. Run installation and execution in a sandbox or container with minimal filesystem access, no unnecessary credentials, restricted network access, and a non-privileged user.
  8. Establish a controlled update process in which new repository commits and dependency-lock changes are reviewed before the pinned version is advanced.
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