Versioned security assessment

Report ID: SA-81733435

6/30/2026, 10:58:32 AM

gpt-image-2 security assessment v2

Skill Security Certification Report

Audit History
Audit model: codex Historical report
Skill name
gpt-image-2
Version
v2
Maintainer
runcomfy-com
Source
github.com/runcomfy-com/skills/tree/main/gpt-image-2/
Coverage
1 Files scanned · 212 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.

Static analysis correctly found external command, network, and filesystem indicators, but most individual hits are Markdown backticks, command examples, URL placeholders, or prose. No prompt injection or confirmed malicious intent was found; the real risk is that the skill directs users to run the RunComfy CLI, send prompts and image URLs to RunComfy endpoints, store or provide a bearer token, and write generated files locally.

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 · 212 Lines analyzed

4 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 7 evidence locations

Filesystem access

May read or write local files.

Observed in 5 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 7 evidence locations

Capability review items (4)
Medium
SKILL.md:24-26SKILL.md:74-78SKILL.md:82-89SKILL.md:93-99SKILL.md:106-109
External RunComfy CLI Invocation
The skill instructs users to install and run the RunComfy CLI, including commands that submit image jobs and download results. This is expected for the skill, but it is a real external command dependency with network and local output effects.
The commands are explicitly documented and use the runcomfy executable. The risk is operational rather than malicious because the examples are static CLI invocations and not hidden code execution.
Medium
SKILL.md:47SKILL.md:207
Bearer Token Use and Local Token Storage
The skill documents RUNCOMFY_TOKEN for CI and states that login stores an API token under the user home configuration directory. This is legitimate credential handling, but marketplace users should understand that a local secret is created or supplied.
The file directly documents token storage and environment-token usage. The stated mode 0600 lowers risk, but credential handling remains sensitive.
Medium
SKILL.md:65-66SKILL.md:195-198SKILL.md:209-210
Remote Processing of Prompts and Image URLs
The skill sends prompts and image URL references to RunComfy model endpoints for remote processing. This is the intended service model, but it has privacy implications for prompts, reference assets, and generated outputs.
The documentation clearly states that API calls and remote URL fetching occur. There is no evidence of unexpected endpoints beyond the documented RunComfy service.
Low
SKILL.md:14-20SKILL.md:86-97SKILL.md:190-195SKILL.md:210
Hardcoded URLs Are Expected Service References
Hardcoded URLs point to RunComfy website, documentation, GitHub, model API endpoints, or example public image URLs. This is expected documentation for a hosted image model skill.
The URLs align with the declared RunComfy integration and model workflow. No unrelated collection endpoint or suspicious third-party callback is present.

Risk findings

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

No confirmed security findings were recorded for this completed audit.

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: codex

Analysis state: Complete

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

Static false positives ignored (3)
Low
SKILL.md:9-10SKILL.md:53-66SKILL.md:147-168
Markdown Backticks Misclassified as Ruby Execution
The external command detections are largely Markdown inline code or fenced command examples, not Ruby shell backtick execution. They still identify a CLI-based workflow, but the specific Ruby execution label is a false positive.
The cited lines are Markdown front matter, tables, and prompt examples. No Ruby source file or executable Ruby syntax is present.
Low
SKILL.md:196-198
URL Placeholder Traversal False Positive
The path traversal detections are URL ellipses such as request status and result paths, not filesystem traversal sequences. No evidence found that the skill reads or writes parent directories through these placeholders.
The suspicious sequences appear inside documented HTTP endpoint paths. They are explanatory placeholders and not local file path operations.
Low
SKILL.md:4SKILL.md:177-179SKILL.md:108
Weak Cryptography and Reconnaissance False Positives
The weak cryptography and system reconnaissance alerts appear to match ordinary prose, front matter, headings, or exit-code documentation. No evidence found of cryptographic code, host reconnaissance, or system inventory behavior.
The lines contain description metadata, an exit-code heading, and a jq example for request IDs. They do not implement cryptography or reconnaissance.

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