Versioned security assessment

Report ID: SA-4AC9DDF8

7/9/2026, 1:06:23 AM

nano-banana-2 security assessment v6

Skill Security Certification Report

Audit History
Audit model: claude Historical report
Skill name
nano-banana-2
Version
v6
Maintainer
agentspace-so
Source
github.com/agentspace-so/runcomfy-agent-skills/tree/main/nano-banana-2/
Coverage
1 Files scanned · 207 Lines analyzed
Policy version
Unavailable

Confirmed finding summary

No confirmed security findings

The completed audit recorded no confirmed security findings. This is not proof that the Skill has no side effects.

Installation context

Historical evidence

This report may not describe the currently installable artifact. Open the current Skill page for install guidance.

Open current Skill page

This report does not block or authorize the manifest or ZIP.

The skill is transparent documentation for using the RunComfy CLI to generate images with Nano Banana 2. Most static findings are Markdown formatting false positives; the remaining confirmed items are the intended external CLI and outbound API behavior, with no evidence of prompt injection, credential exfiltration, or hidden local file access.

Report position

Historical report

Open audit history before using this report to install.

Audit attestation

Not attestable

The required immutable binding is incomplete.

Human verification

Not verified

No human verification is recorded for this report.

Coverage

1 Files scanned · 207 Lines analyzed

8 items shown for review

Limitations

This report does not claim runtime or sandbox execution and does not prove the absence of side effects.

Evidence chain

Follow the evidence from source binding to the install contract. Available evidence supports verification; it is not a safety guarantee.

  1. Source

    Binding unavailable

  2. Artifact

    Identity incomplete

  3. Audit

    Complete

  4. Install contract

    Open manifest to verify

    Open manifest

Capabilities observed

Observed means this report recorded supporting evidence. Not recorded does not prove that a capability is absent.

Contains scripts

May execute code included with the Skill.

Not recorded by this audit

Network access

May connect to external services.

Observed in 4 evidence locations

Filesystem access

May read or write local files.

Observed in 2 evidence locations

Env variables

May read values from the process environment.

Not recorded by this audit

External commands

May invoke commands or programs outside the Skill.

Observed in 40 evidence locations

Capability review items (8)
Medium
SKILL.md:13
Ruby/shell backtick execution
`runcomfy run google/nano-banana-2/text-to-image` through the local
The skill explicitly states that it calls the local RunComfy CLI. This is intended behavior, but it is still real external command execution that submits data to a third-party service.
Medium
SKILL.md:76-80
Ruby/shell backtick execution
```bash
This code block instructs the assistant to run the RunComfy CLI with user prompt input and an output directory. The command is legitimate but represents real local command execution.
Medium
SKILL.md:84-93
Ruby/shell backtick execution
```bash
This batch example invokes the RunComfy CLI and submits prompt data to the remote model service. It is expected functionality, not hidden execution.
Medium
SKILL.md:97-106
Ruby/shell backtick execution
```bash
This final-image example invokes the RunComfy CLI with a JSON body and output directory. It is real external command execution with network effects.
Medium
SKILL.md:110-117
Ruby/shell backtick execution
```bash
This example invokes the RunComfy CLI with web search enabled, causing external command execution and additional remote retrieval by the service. It is documented functionality.
Medium
SKILL.md:198
Ruby/shell backtick execution
The skill invokes `runcomfy run google/nano-banana-2/text-to-image` with a JSON body matching the sc
The text explains that the skill invokes runcomfy and posts to the RunComfy API. This confirms the external command and network behavior is intentional.
Medium
SKILL.md:203-205
Ruby/shell backtick execution
- **Input boundary**: the user prompt is passed as a JSON string to the CLI via `--input`. The CLI d
The line confirms user prompt content is passed into the CLI as JSON. The text says the CLI avoids shell expansion, so injection risk appears low but command execution remains real.
Low
SKILL.md:198
Hardcoded URL
The skill invokes `runcomfy run google/nano-banana-2/text-to-image` with a JSON body matching the sc
The line documents the RunComfy model API endpoint used for request submission. This is intentional outbound network access to a third-party service.

Risk findings

Confirmed security concerns are separated from items that still need review.

No confirmed security findings were recorded for this completed audit.

Remediation

Suggested fixes recorded by this audit. Applying them is the maintainer’s responsibility.

  1. FIX-001
    Medium
    External CLI execution is required for the skill workflow.
    Keep command examples explicit and require normal tool approval before running RunComfy commands.
  2. FIX-002
    Low
    Prompts and generated image requests are sent to RunComfy services.
    Warn users not to submit sensitive prompts or private data unless their RunComfy account and data policy allow it.
  3. FIX-003
    Low
    Authentication depends on a local token file or RUNCOMFY_TOKEN.
    Continue documenting token storage permissions and prefer environment variables in CI or container environments.

Expert evidence

Immutable subject identity, scanner metadata, dismissed matches, and source-level evidence.

Artifact subject

Marketplace commit
Unavailable
Content hash
Unavailable
Tree hash
Unavailable
Skill path
Unavailable
Audit payload hash
Unavailable

Analysis metadata

Audit model: claude

Analysis state: Complete

Scope is limited to the recorded files, lines, methods, and evidence. No runtime or sandbox execution is claimed.

Verify and export

JSON report CITATION.cff SARIF 2.1.0 Signed installation manifest

The manifest and lockfile bind install artifacts to cryptographic hashes. This integrity claim is separate from the security assessment.

Audit attestation: not_attestable