Description

Precision journal monitor using ISSN lookup. No tool-switching allowed.

README (SKILL.md)

Instructions

Name: journal-monitor-xch
Author: chenghan66

You are a highly efficient research assistant. Follow these rules without exception:

Use monitor.py Only:
- Do NOT create your own Python scripts.
- Do NOT use Crossref or internal search unless monitor.py fails after 3 retries.
- The script now supports ISSN lookup, so it is highly accurate.
Self-Translation (MANDATORY):
- Do NOT ask the user to call translation APIs or other skills.
- You have native-level translation capabilities. Use them.
- For every article, translate the English title into professional Chinese immediately.
Output Execution:
- If more than 20 articles are found:
  - Step 1: Tell the user: "Found [count] articles. Processing the full report to your Desktop."
  - Step 2: Use write_file to save ALL data (Bilingual Titles, PMID, Year, Author) to the Desktop.
  - Constraint: Each title must have its Chinese translation right below it.
Template Enforcement:

English: [title] 中文标题: [Your professional Chinese translation] Info: PMID:[pmid] | Year:[year] | Author:[author]

Usage Guidance

This skill appears to do what it claims (ISSN lookup + PubMed monitoring), but several red flags should be resolved before installing: 1) The SKILL.md advertises python3 + Biopython, yet the registry lists no required binaries—install Biopython (pip install biopython) or request the author to fix the metadata. 2) The SKILL.md 'install' entry contains a garbled command ('uv pip install biopython'); treat that as a typo and confirm proper install steps. 3) monitor.py hard-codes Entrez.email to a placeholder—edit the script to set a real contact email (NCBI requires this); do not leave the placeholder. 4) SKILL.md mandates writing a full bilingual report to the Desktop when >20 articles are found—ensure you are comfortable allowing file writes to your Desktop and that the agent's write_file capability is appropriately scoped. 5) Confirm with the author why dependency declarations are missing and ask for a corrected install spec. If you need stronger assurance, ask the publisher to: correct the install metadata, remove the placeholder email or make it configurable via a declared env var, and explicitly declare file-write behavior. These fixes would move the skill from 'suspicious' toward 'benign.'

Capability Analysis

Type: OpenClaw Skill Name: journal-monitor Version: 1.0.11 The journal-monitor skill is a legitimate tool designed to fetch and format research articles from PubMed. The Python script (monitor.py) uses the well-known biopython library to interact with the official NCBI Entrez API, and the instructions in SKILL.md are focused on formatting, translation, and report generation. No evidence of data exfiltration, malicious execution, or harmful prompt injection was found.

Capability Assessment

ℹ Purpose & Capability

The Python script (monitor.py) implements ISSN lookup and PubMed queries via Biopython/Entrez, which aligns with the 'journal monitor' description. However, the registry metadata lists no required binaries or env vars while SKILL.md advertises a dependency on python3 and Biopython (the SKILL.md 'install' line is garbled: 'uv pip install biopython'). This mismatch between declared requirements and actual code is inconsistent and should be clarified.

⚠ Instruction Scope

SKILL.md tightly constrains runtime behavior (use monitor.py only, no other tools unless monitor.py fails after 3 retries) and requires the agent to translate titles to Chinese and to save a full report to the user's Desktop when >20 articles are found. The write_file-to-Desktop instruction implies file system writes; the skill does not declare or justify that requirement in the registry metadata. The mandatory self-translation instruction also forbids use of other translation tools — not malicious by itself but increases the agent's responsibilities and risk if translation is done incorrectly or with sensitive content.

ℹ Install Mechanism

There is no formal install spec in the registry (instruction-only), which is low risk, but SKILL.md includes an 'install' hint recommending python3 and pip-installing Biopython. That is not enforced by the registry and the provided 'uv pip install biopython' looks like a typo/garbled entry. The code itself depends on Biopython (Bio.Entrez), so the missing/incorrect dependency declaration is a correctness concern rather than an obvious supply-chain-malicious indicator.

ℹ Credentials

The skill requests no credentials or environment variables, and the code does not read secrets. One notable artifact: Entrez.email is hard-coded to '[email protected]' with an inline comment asking the user to supply a real email. This is not a secret, but it is required by NCBI policy for reasonable API usage and should not be left as-is. No other env access or unrelated credentials are requested.

✓ Persistence & Privilege

The skill is not always-enabled, has no install script that modifies other skills or system-wide settings, and does not request persistent privileges. The instruction to write files to the Desktop is a local file write but scoped to report output; this is normal for a monitoring tool but should be confirmed with the user.

Version History

v1.0.11

journal-monitor 1.0.11 - Added explicit version field to SKILL.md for better version tracking. - No functional or instruction changes.

v1.0.10

- No code or instruction changes in this release. - Version 1.0.10 is functionally identical to the previous version.

v1.0.9

Version 1.0.9 - Now uses ISSN lookup in monitor.py for improved journal accuracy. - Output template updated: Chinese translation must immediately follow the English title. - For >20 articles, saves all data as a file to Desktop, with both English and Chinese titles for each article. - Instructions clarified to strongly prohibit tool-switching, outside translation services, or extra Python scripting.

v1.0.8

the monitor.py has ben changed.

v1.0.7

- No code or instruction changes detected in this version. - No file changes between 1.0.6 and 1.0.7; behavior and usage remain the same.

v1.0.6

No changes detected in this version. - No file modifications or updates were made. - Skill instructions, description, and metadata remain unchanged.

v1.0.5

journal-monitor 1.0.5 - Refined instructions to enforce strict tool usage—only use data from `monitor.py`, no external or internal searches. - Mandated direct, professional English-to-Chinese translation for every article title. - Standardized output structure in the `.txt` file with required metadata (PMID, year, author) and translation labeling. - Enhanced clarity for large datasets: display article count in chat, delegate detailed content exclusively to the output file. - Strengthened metadata checks; missing fields (e.g., PMID, author) now constitute failure to complete the task.

v1.0.4

**Changelog: journal-monitor v1.0.4** - Introduced a "two-track" output system: top 5 articles shown in chat, full report of all articles saved to file. - Ensured every found article is included in the .txt file; article count and data integrity checks added. - Improved handling of large datasets by splitting file output into chunks if over 15 articles. - Simplified and clarified bilingual title, abstract, and lay summary requirements. - Updated workflow and output formatting for greater reliability and data completeness.

v1.0.3

journal-monitor 1.0.3 - Added a fallback: if no results are found using the journal’s full name, retry the search using the official PubMed Journal Abbreviation. - No other functional or formatting changes.

v1.0.2

- Lay summaries are now limited to a maximum of 100 words (previously ~200 words). - Bilingual titles are now mandatory for all output. - Section inclusion wording clarified: lay summary only included when explicitly requested; abstract omitted for "titles only" requests. - Output formatting instructions streamlined for clarity. - Description and examples updated for better accuracy and consistency.

v1.0.1

journal-monitor 1.0.1 changelog: - Improved output formatting: Each section (title, info, abstract, lay summary) is now clearly separated with newlines and formatted headers for better readability. - Titles are now presented with original English on top, followed by the Chinese translation below. - Added support for user requests for customized report content (e.g., "titles only", skipping abstracts/summaries). - Updated instructions and examples to clarify supported user commands and output structure. - Maintained fuzzy journal recognition and saves report to user’s Desktop.

v1.0.0

- Initial release of journal-monitor skill. - Monitors and summarizes recent articles and reviews (past 7–30 days) from user-specified journals. - Supports fuzzy journal name recognition and correction. - Provides bilingual titles (Chinese/English), original abstracts, and English lay summaries for each article. - Compiles results into a structured text file saved to the user's Desktop. - Notifies the user with a brief list of articles found and file location.

Metadata

Slug journal-monitor

Version 1.0.11

License MIT-0

All-time Installs 0

Active Installs 0

Total Versions 12

Frequently Asked Questions

What is journal-monitor-xch?

Precision journal monitor using ISSN lookup. No tool-switching allowed. It is an AI Agent Skill for Claude Code / OpenClaw, with 165 downloads so far.

How do I install journal-monitor-xch?

Run "/install journal-monitor" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is journal-monitor-xch free?

Yes, journal-monitor-xch is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does journal-monitor-xch support?

journal-monitor-xch is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created journal-monitor-xch?

It is built and maintained by Chenghan66 (@chenghan66); the current version is v1.0.11.

More Skills

journal-monitor-xch