Back to skill

Security audit

nano-banana

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent RunAPI image-generation helper, with disclosed CLI installation and authentication requirements, though users should trust the third-party Homebrew tap before installing it.

Install only if you are comfortable trusting the RunAPI Homebrew tap and CLI with your RunAPI account, selected input media, network requests, and potentially billable task submissions. Use a scoped API key where possible and review CLI updates in sensitive environments.

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:14
Finding
Unpinned Third-Party CLI Dependency## Vulnerability Details **File Location**: `SKILL.md`, lines 14–18 **Vulnerability Type**: Unpinned executable from a third-party Homebrew tap **Risk Level**: Medium ### Vulnerable Code ```yaml install: - kind: brew formula: runapi-ai/tap/runapi bins: - runapi ``` ### Technical Analysis The skill instructs users or agents to install the `runapi` executable from the third-party Homebrew tap `runapi-ai/tap` without pinning a reviewed version, immutable commit, artifact digest, or cryptographic signature. The installed CLI is subsequently trusted to handle authentication, read local media selected for upload, make network requests, and submit potentially billable tasks. Because the dependency is mutable, its effective implementation may change after this skill has been audited. A compromise of the tap, its release infrastructure, or a future package version could introduce attacker-controlled executable code without requiring any change to `SKILL.md`. This finding represents supply-chain exposure. The audited material does not establish that the current package is malicious. ### Attack Path 1. An attacker compromises the third-party Homebrew tap, package release account, or artifact-hosting infrastructure. 2. The attacker publishes a malicious update under the existing `runapi-ai/tap/runapi` formula. 3. An agent follows the skill metadata and installs or upgrades the unpinned formula. 4. Homebrew executes package installation logic and places the compromised `runapi` binary on the system. 5. When the skill invokes the CLI, the compromised binary can operate with the invoking user's privileges and access credentials, local files, and network resources available to that process. ### Impact Assessment Successful exploitation could provide arbitrary code execution with the privileges of the user running Homebrew or the CLI. Within that user's accessible scope, a malicious dependency could capture RunAPI authentication material, read local media suppli ...[truncated 302 chars]
Remediation
## Remediation Suggestions - Pin the CLI to a specifically reviewed version rather than resolving the latest formula dynamically. - Reference an immutable release artifact or source commit and verify its cryptographic digest before installation. - Require verification of official release signatures where the provider supports signed artifacts. - Document the authoritative package repository, publisher identity, and expected release source. - Disable or avoid automatic dependency upgrades in audited environments; review each new version before deployment. - Use narrowly scoped RunAPI credentials with spending limits and rotation procedures. - Execute the CLI with least privilege, restricting filesystem and network access to only what the image-processing task requires. - Consider maintaining a verified internal package mirror so upstream changes cannot enter production without review.
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

Static analysis

No suspicious patterns detected.