Skills/PyPI/mineru-book-mcp

mineru-book-mcp@0.3.0

CAUTION· install only

Publisher claims: "Convert books, PDFs, scans, images, Office files, web pages and ebooks into structured agent skills. Preserves book-to-skill chapter generation, glossary, patterns, cheatsheet, analysis and update workflows, with optional MinerU OCR and layout extraction." Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry PyPI (MCP server)Publisher pypi · publishing for 723 daysInstalls 265Last detonated 3d ago
What would fix thisWatchDispute verdict

Coverage

Nothing executable in the bundle ran; the trace covers the install step only.
Tested
every file read, file written and host contacted by what ran
Not tested
use of the skill by an agent

What this means for the label: Caution here covers what ran. The server's tools were not exercised with a real credential, so nothing about their behavior is asserted; the credential summary below says what to do with that.

Before you give it your token

It will ask for
MINERU_TOKEN or --core-tokensConfiguration, not credentials: BOOK_SKILL_INSTALL_MISSING, BOOK_SKILL_WORKDIR, OBSIDIAN_VAULT, SKILL_SERVICE_DATA
Where it talks to
no vendor host named in the codeAlso mineru.net, calibre-ebook.com, openxmlformats.org, claude.com, ampcode.com, nousresearch.com — not the publisher's domain; check what for before connecting.
What the honeypot saw
no credential files read · no network
What we could not test
use of the skill by an agent

Connect it with a scoped credential, under the guard, and read the third-party hosts first. The code names mineru.net, calibre-ebook.com, openxmlformats.org, claude.com, ampcode.com, nousresearch.com as well as its own vendor. The guard injects MINERU_TOKEN only into requests to the host you bind (and its subdomains) and refuses the rest.

$vet-skills run --net declared -- uvx mineru-book-mcp

Never paste a credential into a chat, a ticket or a report, including ours. How the guard holds it.

Is [email protected] safe to install?

KillerSkills ran it in the KillerSkills honeypot sandbox (1 scenario; nothing executable in the bundle ran; the trace covers the install step only) on 2026-11-11 and scored it 66/100 (Caution). It is not known to be malicious, but it has behaviour or code a security policy should decide on before install. Notable findings: Prompt injection; Hidden zero-width characters detected; Privileged sudo command. Summary: no network, touches nothing outside the project, writes no memory.

Behavior trace

Honeypot run · 0 s · 1 scenarios · 0 deviations
t+CategoryObserved action
0.0sprocessno install scripts in bundle
0.0sprocessscript backend runs setup only
0.0sprocessscript backend runs setup only
0.0sprocessno install scripts to re-run

Static findings

CRITICAL
Prompt injectionskill_service/vendor/SKILL.md:555 · Prompt injection — known jailbreak technique reference… elevated `cmd`, or `New-Item -ItemType Junction` in PowerShell — needs neither Developer Mode nor a symlink privilege; if you attempt it, it does not change the read-back-th…Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
MEDIUM
File type combined with upload actionskill_service/vendor/SKILL.md:149 · File type combined with upload action — agent may exfiltrate local data files…d assets. The standalone MinerU `--json` status and the book-to-skill `metadata.json` must be checked before claiming conversion. A failed or partial run returns nonzero; inspect errors and disclose omitted sources. Never treat extraction …Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
Python subprocess execution detectedskill_service/server.py:19 · Python subprocess execution detectedresult = subprocess.run(cmd, capture_output=True, timeout=timeout, env=env)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
HIGH
Hidden zero-width characters detectedskill_service/vendor/book_to_skill/utils.py:258 · Hidden zero-width characters detected — may conceal content# so "فصل اولویت‌ها" / "فصل اولیه" are not read as chapter 1;Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST07
HIGH
Privileged sudo commandskill_service/vendor/book_to_skill/utils.py:938 · Privileged sudo command — agent may execute with elevated privileges" sudo apt install poppler-utils\n"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Hidden zero-width characters detectedskill_service/vendor/book_to_skill/utils.py:290 · Hidden zero-width characters detected — may conceal contentm["سی‌ام"] = 30 # ZWNJ spelling common in Persian typographyFix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST07
HIGH
Access to sensitive file pathsskill_service/vendor/tools/scan_generated_skill.py:96 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filesr"(?:\.env\b|\bbase64\b|\bsecrets?\b|\bcredentials?\b|\bapi[_ -]?keys?\b)",Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
HIGH
Authorization escalation languageskill_service/vendor/SKILL.md:642 · Authorization escalation language…not create the symlink (a plain copy drifts on the next Update/Fold-in). Enable Developer Mode / create the link manually, or run the skill from ~/.agents/skills" - Hermes Ag…Fix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
CRITICAL
Context shift attemptskill_service/vendor/SKILL.md:642 · Context shift attempt — trying to change operational mode… could not create the symlink (a plain copy drifts on the next Update/Fold-in). Enable Developer Mode / create the link manually, or run the skill from ~/.agents/skills" - Hermes Ag…Fix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
MEDIUM
npx invocation of non-well-known packageskill_service/vendor/SKILL.md:692 · npx invocation of non-well-known package — may execute untrusted codenpx skills add https://github.com/<owner>/<repo_name> --skill <skill_name>Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
HIGH
Collapsed <details> block carrying instructionsskill_service/vendor/docs/how-it-works.md:73 · Collapsed <details> block carrying instructions<summary>Design principles (click to expand)</summary> 1. **Density over completeness** — a 1,000-token summary beats a 10,000-token excerpt 2. **Practitioner vFix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
HIGH
Privileged sudo commandskill_service/vendor/book_to_skill/dependencies.py:29 · Privileged sudo command — agent may execute with elevated privileges"system": [("pdftotext", "poppler-utils", "sudo apt install poppler-utils")],Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Hidden zero-width characters detectedskill_service/vendor/book_to_skill/utils.py:286 · Hidden zero-width characters detected — may conceal content# Fused "بیستم" only — "بیست ام" / "بیست‌ام" are not common spellingsFix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST07
CRITICAL
Prompt injectionskill_service/vendor/SKILL.md:540 · Prompt injection — known jailbreak technique reference# `ln -s` may COPY instead of link (or need Developer Mode / an elevated shell), andFix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
HIGH
Hidden zero-width characters detectedskill_service/vendor/book_to_skill/utils.py:287 · Hidden zero-width characters detected — may conceal content# (unlike "سی ام" / "سی‌ام" for 30), so they stay unmapped on purpose.Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST07
MEDIUM
npx invocation of non-well-known packageskill_service/vendor/SKILL.md:659 · npx invocation of non-well-known package — may execute untrusted code> "Want me to publish this skill to GitHub so any Agent Skills host can install it with `npx skills add`? (yes / skip)"Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
HIGH
Authorization escalation languageskill_service/vendor/SKILL.md:540 · Authorization escalation language# `ln -s` may COPY instead of link (or need Developer Mode / an elevated shell), andFix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
HIGH
Hidden zero-width characters detectedskill_service/vendor/tools/validate_skill.py:152 · Hidden zero-width characters detected — may conceal content# leading  makes text.startswith("---") false and frontmatter is missed.Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST07
CRITICAL
HTML comment carrying imperative instructions to the agentskill_service/vendor/SKILL.md:391 · HTML comment carrying imperative instructions to the agentReproduce or reconstruct one concrete example the author works through: a sample document, a dialogue, a filled-in template, a before/after, or a decision walkeFix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
HIGH
Authorization escalation languageskill_service/vendor/SKILL.md:555 · Authorization escalation language… elevated `cmd`, or `New-Item -ItemType Junction` in PowerShell — needs neither Developer Mode nor a symlink privilege; if you attempt it, it does not change the read-back-th…Fix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
CRITICAL
Prompt injectionskill_service/vendor/SKILL.md:642 · Prompt injection — known jailbreak technique reference…not create the symlink (a plain copy drifts on the next Update/Fold-in). Enable Developer Mode / create the link manually, or run the skill from ~/.agents/skills" - Hermes Ag…Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
MEDIUM
Python subprocess execution detectedskill_service/vendor/book_to_skill/parsers/calibre.py:16 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedskill_service/vendor/book_to_skill/parsers/pdf.py:102 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
File type combined with upload actionskill_service/vendor/SKILL.md:159 · File type combined with upload action — agent may exfiltrate local data filesRead this run's `metadata.json` (the `Meta ->` path from the extraction output) and present the user with an estimate **before doing any generation**:Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
Python subprocess execution detectedskill_service/vendor/book_to_skill/parsers/pdf.py:81 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedskill_service/vendor/tools/evals/manifest.py:113 · Python subprocess execution detectedresult = subprocess.run(Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST05
MEDIUM
Python subprocess execution detectedskill_service/vendor/book_to_skill/dependencies.py:124 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
File type combined with upload actionskill_service/vendor/tools/evals/manifest.py:4 · File type combined with upload action — agent may exfiltrate local data filesUsage: python3 tools/evals/manifest.py config.json [--output manifest.json]Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST07
MEDIUM
File permission modification commandskill_service/vendor/book_to_skill/config.py:26 · File permission modification command — agent changes file ownership/permissions# would then populate and chmod to 0700.Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
MEDIUM
Python subprocess execution detectedskill_service/vendor/book_to_skill/parsers/pdf.py:177 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
File type combined with upload actionskill_service/vendor/SKILL.md:718 · File type combined with upload action — agent may exfiltrate local data filesAnalyze the new extracted text in this run's `full_text.txt` (the `Text ->` path from the extraction output) to identify if the new content represents:Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07

Cost to the agent

Instructions loaded
≈11,160 tok
1 file · 45 in bundle

Tokens estimated at four characters each. Skills load their instruction text when triggered; the figure is what that costs.

Known vulnerabilities

No advisories in OSV for mineru-book-mcp@0.3.0 as of 3d ago. This is the package itself, not its dependencies.

What would make this Verified

14 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%
  1. Unclaimed publisher≈ +8 on the totalClaim pypi: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
  2. Hidden instructions≈ +6 on the totalRemove HTML comments, zero-width characters and white-on-white text that carry instructions. Everything the agent is meant to read should be visible to the person reading the file.
  3. Hidden zero-width characters detected≈ +4 on the totalKeep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
  4. Privileged sudo command≈ +4 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
  5. Access to sensitive file paths≈ +4 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
  6. File type combined with upload action≈ +2 on the totalRemove the call, or declare the host in frontmatter and explain in the README what is sent and why.
Publish a new version to PyPI, then queue it here. Not your skill? Claim the handle to get verdict changes by email.Claim pypiPropose a patch to the maintainer