Skills skywork-document
📦
skywork-document
Content revision r1 High Risk 🌐 Network access🔑 Env variables📁 Filesystem access⚙️ External commands
Create Skywork Documents from Files
Long document requests often need structure, reference files, and formatted delivery. This skill uses Skywork document services to create polished files from prompts and uploaded materials.
Supports: Claude Codex Code(CC)
Stop for confirmation before installing.
Review the plan and obtain explicit user consent before changing files.
Install with my Agent
Copy this request to your Agent. It includes the canonical Skill page and manifest.
Agent request
Review the Skillstore skill "skywork-document" from https://skillstore.io/skills/skyworkai-skywork-document.md and its manifest at https://skillstore.io/api/skills/skyworkai-skywork-document/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.Your Agent should still show its plan and request any confirmation required by the security policy.
Agent-readable resources
Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.
Test it
Using "skywork-document". User asks for a DOCX proposal based on a short business brief.
Expected outcome:
A downloadable DOCX proposal with sections for goals, scope, timeline, pricing notes, and recommended next steps.
Using "skywork-document". User provides a PDF research paper and asks for a summary report.
Expected outcome:
A PDF report with an executive summary, major findings, evidence notes, limitations, and a local saved file path.
Using "skywork-document". User asks for a Markdown project plan in German from several source files.
Expected outcome:
A Markdown document with milestones, responsibilities, risks, assumptions, and references derived from the uploaded materials.
Security Audit
High Riskv5 • 7/7/2026 Open versioned report
The skill is a cloud document generator with real network, file upload, credential storage, and browser-login behavior. I confirmed the substantive risks around external processing, token caching, endpoint overrides, and Windows shell use, while marking Markdown backtick detections as false positives. No prompt injection attempt was found in the reviewed files.
4
Files scanned
923
Lines analyzed
21
Review items
0
False positives ignored
Confirmed security concerns (3)
High
scripts/parse_file.py:40-41scripts/parse_file.py:85-91scripts/create_doc.py:65-85scripts/create_doc.py:95
External Processing of User Files and Prompts
The skill reads local reference files, uploads them to Skywork, and sends document prompts to Skywork APIs. This is expected functionality, but users must understand that their content leaves the local environment.
The code directly reads file bytes, posts them to a remote API, and sends document content in the request payload. The behavior is central to the skill and clearly evidenced.
High
Credential Cache Lacks Permission Hardening
The authentication module saves Skywork token data to ~/.skywork_token using a normal file write. No chmod, secure file creation, or platform credential store is used around the token cache.
The token cache path and write operation are explicit, and there is no nearby permission hardening. This makes the credential storage concern high confidence.
Medium
Environment-Controlled Endpoints Can Redirect Sensitive Data
Several API base URLs can be overridden by environment variables before files, prompts, or token checks are sent. This is useful for configuration but risky in shared or compromised environments.
The endpoint overrides are explicit and are used by network calls that process sensitive data. The risk depends on environment trust, so confidence is high but not absolute.
Capability review items (21)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
High
Python subprocess.Popen
subprocess.Popen(["start", url], shell=True, start_new_session=True,
The Windows branch uses subprocess.Popen with shell=True to open a login URL. The URL is partly environment-derived, so shell interpretation creates command-injection risk.
High
Hidden file in home directory
Token is stored in user home directory as ~/.skywork_token (JSON format).
The module documentation states that the Skywork token is stored in ~/.skywork_token. Persisting an authentication token in a hidden home file is a real credential-storage risk.
High
Hidden file in home directory
TOKEN_FILE = os.path.expanduser("~/.skywork_token")
TOKEN_FILE is explicitly set to ~/.skywork_token. This stores a reusable authentication token in a hidden file under the user home directory.
High
Python file write/append
with open(TOKEN_FILE, "w", encoding="utf-8") as f:
The code writes token data to the token cache file. Because the file stores credentials, the write is security sensitive and should be permission-hardened.
High
Hidden file in home directory
2. Cached token file `~/.skywork_token` → validate via API, if valid, use it
The skill instructions explicitly tell the agent to use a cached token file at ~/.skywork_token. That documents and enables persistent credential storage in a hidden home file.
Medium
Python file write/append
with open(args.output, "w", encoding="utf-8") as f:
The script writes parsed content to a user-supplied output path when requested. This is useful but can overwrite local files if the caller chooses an unsafe path.
Medium
Hidden file access
Token is stored in user home directory as ~/.skywork_token (JSON format).
The documentation describes reading and writing a hidden token file. That file contains credential material, so hidden-file access is security relevant.
Medium
Hidden file access
TOKEN_FILE = os.path.expanduser("~/.skywork_token")
The code defines ~/.skywork_token as the token cache path. Access to this hidden file can expose or modify an authentication token.
Medium
Python environment access
env_token = os.environ.get("SKYBOT_TOKEN", "")
The code reads SKYBOT_TOKEN from the environment and returns it as the credential. Environment token access is legitimate but handles sensitive authentication material.
Medium
Python environment access
env_token = os.environ.get("SKYBOT_TOKEN", "")
The command-line path reads SKYBOT_TOKEN and validates it through the Skywork API. This is legitimate authentication behavior but still accesses a sensitive environment token.
Medium
Hidden file access
2. Cached token file `~/.skywork_token` → validate via API, if valid, use it
The instructions direct use of ~/.skywork_token as part of authentication. The referenced hidden file contains credential data, so access to it is security relevant.
Low
Python HTTP libraries
from urllib.request import Request, urlopen, urlretrieve
The imported urllib functions are used to call the Skywork document API and download generated files. This is intentional network behavior that sends document prompts and receives remote file URLs.
Low
Hardcoded URL
base_url = os.environ.get("SKYWORK_GATEWAY_URL", "https://api-tools.skywork.ai/theme-gateway").rstri
The default gateway URL points to a production Skywork endpoint used for document creation. This is a real external network dependency for user content.
Low
Python environment access
base_url = os.environ.get("SKYWORK_GATEWAY_URL", "https://api-tools.skywork.ai/theme-gateway").rstri
SKYWORK_GATEWAY_URL can override the endpoint that receives document prompts and file references. This is legitimate configuration but can redirect sensitive data if the environment is untrusted.
Low
Python HTTP libraries
from urllib.request import Request, urlopen
The imported urllib functions are used to upload reference files to the Skywork parse API. That creates a real external data transfer path.
Low
Hardcoded URL
base_url = os.environ.get("SKYWORK_GATEWAY_URL", "https://api-tools.skywork.ai/theme-gateway").rstri
The default gateway URL is the production Skywork endpoint used for parsing uploaded files. This is a real network dependency for user-provided documents.
Low
Python environment access
base_url = os.environ.get("SKYWORK_GATEWAY_URL", "https://api-tools.skywork.ai/theme-gateway").rstri
SKYWORK_GATEWAY_URL can override the endpoint that receives uploaded reference files. This is legitimate configuration but can redirect sensitive files if the environment is untrusted.
Low
Hardcoded URL
API_BASE = os.environ.get("SKYWORK_API_BASE", "https://api.skywork.ai")
The default API base points to Skywork and is used for login key retrieval and token validation. This is a real external authentication dependency.
Low
Hardcoded URL
LOGIN_BASE_URL = os.environ.get("SKYWORK_WEB_BASE", "https://skywork.ai")
The default web base is used to build browser login URLs. This is legitimate but still causes an external authentication flow.
Low
Python environment access
API_BASE = os.environ.get("SKYWORK_API_BASE", "https://api.skywork.ai")
SKYWORK_API_BASE can override the authentication API endpoint. This is configuration, but it can redirect token validation traffic in an untrusted environment.
Low
Python environment access
LOGIN_BASE_URL = os.environ.get("SKYWORK_WEB_BASE", "https://skywork.ai")
SKYWORK_WEB_BASE can override the login URL base. This can direct users to a different login page if the environment is manipulated.
Risk Factors
🌐 Network access (9)
🔑 Env variables (7)
📁 Filesystem access (8)
⚙️ External commands (59)
scripts/skywork_auth.py:87 scripts/skywork_auth.py:90 scripts/skywork_auth.py:93 SKILL.md:21-24 SKILL.md:24-27 SKILL.md:27-28 SKILL.md:28-31 SKILL.md:31-49 SKILL.md:49 SKILL.md:50-53 SKILL.md:53 SKILL.md:54-55 SKILL.md:55 SKILL.md:58-66 SKILL.md:66-68 SKILL.md:68-71 SKILL.md:71-73 SKILL.md:73-76 SKILL.md:76-85 SKILL.md:85-88 SKILL.md:88-95 SKILL.md:95-97 SKILL.md:97-98 SKILL.md:98-104 SKILL.md:104-106 SKILL.md:106 SKILL.md:111-117 SKILL.md:117-120 SKILL.md:120-127 SKILL.md:127-129 SKILL.md:129-132 SKILL.md:132-140 SKILL.md:140-144 SKILL.md:144-146 SKILL.md:146-147 SKILL.md:147-153 SKILL.md:153-155 SKILL.md:155 SKILL.md:162-163 SKILL.md:163-165 SKILL.md:165-168 SKILL.md:168-169 SKILL.md:169-172 SKILL.md:172-173 SKILL.md:173-174 SKILL.md:174 SKILL.md:175 SKILL.md:176 SKILL.md:177 SKILL.md:185 SKILL.md:186 SKILL.md:189-191 SKILL.md:191 SKILL.md:199 SKILL.md:200-201 SKILL.md:201-202 SKILL.md:202-209 SKILL.md:209-210 SKILL.md:210-213
Audited by: codex View Audit History →
Share & cite this report
Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.
Copy report link
https://skillstore.io/skills/skyworkai-skywork-document/audits/5?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown badge
[](https://skillstore.io/skills/skyworkai-skywork-document?utm_source=security_passport_badge)HTML badge
<a href="https://skillstore.io/skills/skyworkai-skywork-document?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/skyworkai-skywork-document/security.svg" alt="Skillstore security assessment" loading="lazy"></a>Embed card
<iframe src="https://skillstore.io/embed/skills/skyworkai-skywork-document.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>Academic citations (APA · BibTeX · CFF)
APA citation
SkyworkAI. (2026). skywork-document security audit report (audit version 5) [Author version unspecified]. Skillstore. https://skillstore.io/skills/skyworkai-skywork-document/audits/5BibTeX citation
@techreport{skyworkai-skyworkai-skywork-document-2026,
author = {SkyworkAI},
title = {skywork-document security audit report (audit version 5)},
institution = {Skillstore},
year = {2026},
number = {5},
url = {https://skillstore.io/skills/skyworkai-skywork-document/audits/5},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "skywork-document security audit report (audit version 5)"
version: "unspecified"
type: report
authors:
- name: "SkyworkAI"
date-released: "2026-07-07"
url: "https://skillstore.io/skills/skyworkai-skywork-document/audits/5"
identifiers:
- type: other
value: "skillstore:skyworkai-skywork-document:audit:5"
description: "Skillstore immutable audit report identifier"
Skillstore Score
Why this score Evidence Confidence: Medium 45
Architecture
85
Maintainability
87
Content
69
Community
83
Spec Compliance
What You Can Build
Prepare Client Reports
Create structured reports from a prompt and supporting PDFs, spreadsheets, or notes.
Convert Research Into Documents
Turn gathered research and source files into a formatted DOCX, PDF, HTML, or Markdown document.
Draft Business Materials
Generate polished plans, briefs, summaries, and proposals in the requested language and format.
Try These Prompts
Create a Simple Document
Create a three-page DOCX overview about renewable energy trends in English with clear headings and a concise conclusion.
Use One Reference File
Use the attached PDF as the source and create a client-ready PDF summary with key findings, risks, and next steps.
Combine Multiple Sources
Create a market analysis report from the attached files. Compare sources, highlight conflicts, and include an executive summary.
Produce a Multilingual Formal Report
Generate a formal Markdown report in Spanish using the uploaded research. Include methodology, findings, recommendations, and appendix notes.
Best Practices
- State the desired format, language, audience, and document structure before generation starts.
- Review files for sensitive information before uploading them to Skywork cloud services.
- Use clear titles without spaces when passing titles to the document creation script.
Avoid
- Do not use this skill for short answers that do not need a generated file.
- Do not upload confidential files unless the user accepts external processing by Skywork.
- Do not pass vague prompts when the document needs a specific structure or tone.
Frequently Asked Questions
Does this skill require a Skywork account?
Yes. It needs a valid Skywork token from the environment, token cache, or browser login flow.
Which output formats are supported?
The scripts support DOCX, PDF, HTML, and Markdown output formats.
Can it use uploaded files as sources?
Yes. It can parse files through Skywork and pass the returned file IDs into document generation.
Where are generated files saved?
Generated files are downloaded to the current working directory when a file URL is available.
Does it send data to an external service?
Yes. Prompts, uploaded files, and document requests are processed by Skywork cloud APIs.
Can it create documents in languages other than English?
Yes. The script accepts a language option, including English, Chinese, Japanese, Korean, French, German, and Spanish.
Developer Details
Author
SkyworkAILicense
MIT
Skillstore revision
r1
Version notice
The author did not declare a version.
Ref
f93e9bb0daca99badb6a7e574b97737155d57cb3
Maintenance freshness
7/18/2026
Usage
9 downloads · 152 views
File structure