This is a portfolio-facing Agent/Plugin project for journal targeting, source verification, and research workflow decisions. It is part of my reusable academic-agent toolkit for Claude Code and Codex.
Two agent plugins for choosing target journals before a paper is written.
This repository packages two independent academic journal research plugins:
journal-research-cn: Chinese journal targeting with domestic catalogs, rank labels, Muchong-style author feedback, and official identity checks.journal-research-en: English and SCI journal targeting with OpenAlex, Crossref, dblp, B!SON, arXiv, Google Scholar style exploration, LetPub profile checks, authority verification, and predatory-risk screening.
The core workflow is practical: build a candidate pool, verify the sources, screen risk, then output a decision matrix with 2 primary journals, 2 backups, and 1 not-recommended target. The plugins are separated because Chinese and English journal data sources, anti-bot boundaries, and ranking systems are materially different.
Many students and early researchers ask an AI model to recommend a journal after the paper is nearly done. That is backwards. Journal targeting should happen before the experiment and writing plan are locked, because venue choice affects methods, formatting, timeline, risk, and review difficulty.
These plugins turn that judgment into an agent-readable workflow. They do not claim to replace supervisor review or editorial judgment. They make the evidence gathering repeatable.
plugins/
journal-research-cn/
skills/
cn-orchestrator/
cn-discover-by-catalog/
cn-discover-by-paper/
easyscholar-rank/
muchong-cn-journal-review/
nppa-validator/
journal-research-en/
skills/
journal-research-orchestrator/
ai-journal-match/
openalex-explore/
crossref-validator/
dblp-cs-explore/
bison-oa-explore/
arxiv-preprint-explore/
google-scholar-explore/
letpub-sci-journal-review/
journal-authority-check/
predatory-risk-check/
Each plugin can be installed, read, and removed independently. The two plugins can refer to each other as workflow neighbors, but they do not require a shared hidden runtime state.
git clone https://github.com/kkunkunya/journal-research-agent-plugins.git
cd journal-research-agent-plugins
cp examples/local-config.example.txt local-config.txt
# Fill your own keys in local-config.txt, then export only what you need.
claude plugin install ./plugins/journal-research-en --scope local
claude plugin install ./plugins/journal-research-cn --scope localFor Codex, OpenClaw, Hermes, and crawler-style agents, see:
Chinese documentation:
This public repository does not ship maintainer-owned API keys, tokens, cookies, or private endpoints. Customers configure their own provider credentials locally.
Supported local key names:
BAIDU_API_KEY: Baidu Scholar style search adapters for Chinese paper-reverse discovery.EASYSCHOLAR_SECRET_KEY: easyScholar rank profile adapter.OPENALEX_API_KEY: optional OpenAlex polite pool key; anonymous access is used when available.FIRECRAWL_API_KEY: optional crawler fallback for pages that need a crawler service.
Use examples/local-config.example.txt as the template. Do not commit your filled local config.
This package is designed for:
- Claude Code plugin installation.
- Codex local marketplace registration and project-session enablement.
- OpenClaw or Hermes agents that can read a GitHub repository URL but cannot install plugins directly.
- Generic crawler agents that must emulate the skill by reading
SKILL.mdfiles and references.
Crawler agents should start from docs/AGENT_CRAWLER_GUIDE.md.
agent-plugin, ai-agents, claude-code, codex, openclaw, hermes, academic-writing, journal-selection, literature-research, research-workflow
MIT. See LICENSE.