Security audit
Security checks for vulnerabilities and agentic risk
This skill is a straightforward Recraft API wrapper for image generation and editing, with one documented account-info command users should run knowingly.
Install only if you are comfortable giving the skill a Recraft API token and sending prompts or selected input images to Recraft. Avoid running user-info unless you want account details such as email and credits printed into the session transcript or logs.
The skill description presents the functionality as image generation/editing, but the documented commands also include a user-info operation that retrieves account metadata including email and credits. Undeclared data-access behavior is dangerous because users may authorize the skill expecting only image processing, while the skill can also expose personal/account information via network/API calls.
The skill invokes a script that uses network access, reads an API token from the environment, and writes output files, but the manifest does not declare an explicit tool scope such as permissions or allowed-tools. This weakens least-privilege controls and makes it harder for reviewers and runtime policy to understand or constrain what the skill can do.
The documented usage includes a user-info command that retrieves account data, which is outside the manifest’s stated image-processing scope. This creates a scope mismatch that can surprise users and unnecessarily exposes personal/account information through a skill that should only manipulate images.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
response = requests.post(url, headers=headers, files=files, data=data)
else:
headers["Content-Type"] = "application/json"
response = requests.post(url, headers=headers, json=json_data)
if not response.ok:
print(f"Error: API request failed with status {response.status_code}", file=sys.stderr)
The cmd_user_info function calls /users/me and prints account attributes such as ID, name, email, and credits. For an image-processing skill, this is an unnecessary data-access capability that broadens exposure of sensitive profile data and may leak it to logs, transcripts, or downstream tooling.
Retrieving account profile information is not justified by the stated purpose of generating and editing images. This violates least privilege by using the bearer token for unrelated profile access, increasing the blast radius if the skill is invoked in automated or delegated contexts.
No suspicious patterns detected.