Security Audit

openai/skills:skills/.system/openai-docs

github.com/openai/skills

AI SkillCommit 49f948faa925

CAUTION

Scanned about 9 hours ago

Critical

Immediate action required

High

Priority fixes suggested

Medium

Best practices review

Low

Acknowledged / Tracked

Trust Assessment

openai/skills:skills/.system/openai-docs received a trust score of 73/100, placing it in the Caution category. This skill has some security considerations that users should review before deployment.

SkillShield's automated analysis identified 2 findings: 0 critical, 2 high, 0 medium, and 0 low severity. Key findings include Unsafe deserialization / dynamic eval, Command Injection via Unsanitized Arguments in resolve-latest-model-info.js.

The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. All layers scored 70 or above, reflecting consistent security practices.

Last analyzed on July 17, 2026 (commit 49f948fa). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.

Layer Breakdown

Manifest Analysis

85%

Static Code Analysis

100%

Dependency Graph

100%

LLM Behavioral Safety

85%

Behavioral Risk Signals

Network Access

1 finding

Filesystem Write

1 finding

Shell Execution

2 findings

Dynamic Code

2 findings

Security Findings2

Severity	Finding	Layer	Location
HIGH	Unsafe deserialization / dynamic eval Decryption followed by code execution Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions.	Manifest	skills/.system/openai-docs/scripts/resolve-latest-model-info.js:7
HIGH	Command Injection via Unsanitized Arguments in resolve-latest-model-info.js The script `scripts/resolve-latest-model-info.js` accepts a `--source` or `--url` argument. If this argument does not start with 'file://' or 'http(s)://', it resolves the path using `path.resolve(source)` and reads it via `fs.readFile`. However, if the source is a URL, it performs a `fetch(source)`. An attacker who can manipulate the arguments passed to this script (for example, via prompt injection manipulating the tool execution arguments) can force the agent to fetch arbitrary external URLs or read arbitrary local files, leading to Server-Side Request Forgery (SSRF) or Local File Inclusion (LFI). Restrict the allowed protocols and domains for the `--source` argument. Validate that the URL points to an official, trusted OpenAI domain (e.g., developers.openai.com) and prevent arbitrary local file paths from being read unless they are explicitly whitelisted.	LLM	scripts/resolve-latest-model-info.js:34

Scan History

Embed Code

[![SkillShield](https://skillshield.io/api/v1/badge/e2ca4655c926de03.svg)](https://skillshield.io/report/e2ca4655c926de03)