AI Research Skills

17 portable SKILL.md research skills for Claude Code, Codex, Cursor and other AI clients, covering an eight-stage lifecycle from literature discovery to manuscript release.

Stars
★ 667
Last updated
1d ago
License
MIT
Primary language
Python

At a glance

How it runs
Agent plugin / skillLibrary / SDK
Works with
Universal · cross-platformClaude CodeCodex · Claude API (Partial support)
Cost
Free software; you pay for model usage
Setup effort
Medium · a few setup steps
You'll need
Python (research-hub-pipeline)Zotero desktop (optional, local API)codex CLI (for codex-delegate)agy CLI (for antigravity-delegate)Shell / CLINetwork accessLocal filesystem
Typical use
A grad student who has read 50 papers but can't tell which gap is thesis-worthy: use gap-to-topic for a three-gate decision dossier.
Not a fit if
  • Users wanting a zero-config turnkey product; teams needing corpus-scale validation or vendor support; researchers outside water resources / agent-based modeling
Maintenance
Commits in the last 30 dayslast commit 2026-11-16latest release v1.7.18 (2026-10-29)+3.4 ★/day over 7 days
Source review
72/100 · Some gaps

What does this agent do, and when should you use it?

AI Research Skills is a portable SKILL.md catalog (registry) that structures research into eight stages from literature discovery to manuscript release, with 17 skills and 5 installable plugins. The core asset is the portable SKILL.md instruction file; the Claude Code marketplace is just a convenience layer, and Codex CLI, Cursor, Gemini CLI, Hermes, and generic API clients can all load the same SKILL.md. An optional runtime, research-hub (pip install research-hub-pipeline), provides executable, resumable workflows, while an optional agent-collab-harness adds policy/checkpoint governance. The catalog rests on three design principles: manifests (.research/, .paper/) that persist research context across AI sessions, YAML schemas with anti-leakage rules governing cross-skill handoffs, and character-driven routing that delegates mechanical work to Codex/Antigravity. Skills integrate with Zotero, Obsidian, and NotebookLM, all three of which are optional inside research-hub. Automation pauses at five explicit human gates; read-only and reversible steps proceed automatically.

Once installed, skills trigger from natural language: literature-triage-matrix builds a matrix comparing papers by method, data, claim, and limitation; gap-to-topic produces a topic_dossier.gaps.yml through three go/no-go gates (is the gap open? a real contribution? feasible?); research-design-helper turns the gap into design_brief.md via Socratic dialog; research-context-compressor writes .research/ manifests so future AI sessions skip the rescan; paper-memory-builder generates .paper/claims.yml with a status enum and anti-leakage rule—a claim with empty evidence_artifacts is forced to status: gap with a gap_reason. research-workflow-orchestrator is a resumable end-to-end controller coordinating orient→scope→discover→synthesize→design→execute→write→release. research-hub (a Python CLI) chains arXiv/Semantic Scholar search, Zotero ingest, Obsidian note sync, and NotebookLM brief verification. codex-delegate and antigravity-delegate hand Claude's work to Codex CLI / Antigravity CLI, with research-hub-multi-ai writing a coordination file each delegate reads instead of the parent context. Every stage's output is the next stage's structured input; handoffs are documented as YAML schemas, and downstream skills refuse malformed handoffs (e.g. status: gap claims without gap_reason are rejected at Stage 7).

  1. A grad student who has read 50 papers but can't tell which gap is thesis-worthy: use gap-to-topic for a three-gate decision dossier.
  2. A PhD candidate building a comparative literature review: use research-hub for search/ingest, then literature-triage-matrix for a method/data/limitation matrix.
  3. A manuscript author worried about AI-sounding prose and hallucinated citations: use paper-memory-builder's claims.yml anti-leakage rule plus academic-writing-skills for terminology, repetition, and overclaim audits.
  4. A multi-tool user designing with Claude, scaffolding code with Codex, and finishing mechanical work in Antigravity: use research-hub-multi-ai for character-driven routing with coordination files.
  5. Research support staff already running Zotero/Obsidian/NotebookLM: install zotero-skills and research-hub to unify them into one library-management loop.
  6. Agent/harness builders: reuse the architecture, state, HITL, and recovery design documented in docs/for-agent-harness-builders.md.

Key setup facts

Pulled from the install and usage notes below. The project README has the current values.

Install / run
  • pip install research-hub-pipeline

How do you install or deploy this agent?

Fastest path (Claude Code, ~30 seconds):

claude plugin marketplace add WenyuChiou/ai-research-skills
claude plugin install research-workspace@ai-research-skills
claude plugin list

Add optional plugins as needed:

claude plugin install academic-writing-skills@ai-research-skills
claude plugin install zotero-skills@ai-research-skills
claude plugin install codex-delegate@ai-research-skills
claude plugin install antigravity-delegate@ai-research-skills

Install all 5 plugins in one go:

bash scripts/install-all.sh

Optional Python runtime:

pip install research-hub-pipeline
research-hub setup --persona researcher
research-hub doctor

For non-Claude-Code hosts (Codex CLI, Cursor, Windsurf, Gemini CLI, generic API clients): load the repo's SKILL.md into the host's skills/rules directory or inline it into the prompt. Hermes 0.13.0 is verified end to end for literature-triage-matrix; OpenClaw is structurally compatible but not release-grade verified.

How do you use this agent?

After installing, describe your goal in natural language in Claude Code, e.g.:

Build a literature triage matrix for human-in-the-loop research agents. Verify identifiers, preserve contradictions, and show me the evidence gaps before recommending next steps.

Or name the skill explicitly:

Use `literature-triage-matrix` to compare these 5 papers.

With the Python runtime, run the automated literature pipeline:

research-hub auto "agent-based modeling" --max-papers 3 --no-nlm

Cautions: enable Zotero's local API in the desktop app before any write operations, and always export a Zotero backup first (File → Export Library → Zotero RDF)—zotero-library-curator is read-only, but zotero-skills can merge, delete, and rebind items. Chinese-language literature and Chinese output are both supported.

What are this agent's strengths and limitations?

Pros
  • The 17 skills hand off via YAML schemas rather than free text, with status enums and anti-leakage rules that structurally block evidence-free claims from shipping as supported.
  • Portability is audited: the universal SKILL.md layer scored 11/14 zero-edit portable in the 2026-05-10 audit, with a recorded Hermes 0.13.0 end-to-end install verification.
  • Five explicit human gates are built into the correctness boundary—destructive actions (external writes, release authorization) pause while read-only reversible work proceeds automatically.
  • The maintainer publishes per-skill time/turn expectation ranges and real dogfood artifacts (test-corpus), making cost estimation concrete.
Limitations
  • Built and tested by one graduate-student researcher; not corpus-scale validated, with domain bias toward water resources and agent-based modeling—unvalidated for social sciences, ML, or clinical writing.
  • A Claude Code marketplace install guarantees nothing about other hosts' live behavior; OpenClaw is not release-verified, and some hosts need cosmetic <skill-root> path edits.
  • Full capability is spread across multiple source repos (research-hub, academic-writing-skills, zotero-skills, etc.); this catalog is only an index, so debugging crosses repos.
  • CI performs no claude plugin install round-trip; the actual install and trigger path is verified by the maintainer between releases.

How does this agent compare with similar options?

The README situates the project within the agentic AI learning roadmap (awesome-agentic-ai-zh) but names no direct competitor; unlike prompt-template collections, it enforces inspectable YAML-schema handoffs and human gates, with skill sources split across standalone repos.

Key facts side by side with the most closely related agents.

Agent Source review Form / cost Stars Updated Language Full support on
AI Research Skills This agent 72 · Some gaps Agent plugin / skillFree + model costs ★ 667 1d ago Python Claude Code
LLM Agents Papers 57 · Major gaps Library / SDKFree ★ 3.1k 3d ago — —
COMPASS Skills 72 · Some gaps CLIFree ★ 770 2d ago Python Codex · Claude Code
Obsidian Second Brain 63 · Some gaps Agent plugin / skillFree + model costs ★ 5.4k 1d ago Python Codex · Claude Code

How does FollowAgents rate this agent?

FollowAgents source review · FARS-2.1
Some gaps
72/ 100 5-point scale 3.6 / 5
Trust 17/29
Reliability 9/14
Adaptability 15/18
Convention 15/18
Effectiveness 10/13
Verifiability 6/8
Why each dimension lost points
Trust17 / 29 · 2.9/5

The files describe human decision gates, require appropriate authorization for Zotero and external writes, data costs, and model calls, and distinguish local instructions from connected operations. Deductions: the main skill content resides in upstream repositories and is not included in this packet, so full permission controls and sensitive-data handling cannot be confirmed; rollback is mentioned, including per-paper rollback, but implementation detail is limited.

Reliability9 / 14 · 3.2/5

The documentation distinguishes skills, the Python runtime, and host capabilities, and provides `doctor`, install checks, and troubleshooting guidance; the CI and test samples show checks for catalog and source contracts. Deductions: this evidence cannot establish consistency across all upstream skills or runtime availability for the reviewed revision; failure reporting is described, but the packet does not contain enough product content to assess all failure paths.

Adaptability15 / 18 · 4.2/5

The intended users include researchers, support staff, and builders, with workflows for literature, writing, and research planning. The docs distinguish hosts, optional integrations, prompt-only skills, and operations requiring the Hub CLI, and identify hosts such as OpenClaw as not yet release-grade verified. Deductions: cross-host loading and trigger precision depend on each host, and only partial verification is provided here.

Convention15 / 18 · 4.2/5

The README, install guide, directory, FAQ, examples, and bilingual docs form a clear information structure; the MIT license has explicit text, and prerequisites, limitations, and version information are documented. Deductions: the packet lacks a complete changelog and a clear maintenance-responsibility or reporting path; external skill sources use floating branch references.

Effectiveness10 / 13 · 3.8/5

The docs describe cross-stage handoffs, structured artifacts, and explicit human gates, and give usable install and prompt examples across several research tasks. Deductions: benefit and time estimates are maintainer observations rather than benchmarks, and this packet cannot confirm that upstream skills not included here reliably produce the described outputs.

Verifiability6 / 8 · 3.8/5

The docs distinguish maintainer observations, host-specific verification, unverified compatibility, and runtime contracts; the test samples also state their evidence limits rather than treating fixture passes as host verification. Deductions: several product capabilities are supported only by README claims, and the packet lacks upstream SKILL.md files, complete immutable source pins, and independent corroboration.

Risks and how to mitigate them
  • This assessment uses only the supplied static materials. Upstream SKILL.md files and other files not provided are unknown; no installation or workflow was run.
  • External skill sources refer to floating branches such as `master` or `main`; the install guide says Claude Code marketplace verification does not establish that other hosts loaded the skills.
  • Hub pipeline operations, Zotero CRUD, model calls, and other connected actions may involve data transfer, cost, or writes; confirm the required authorization and destination before performing them.
Evidence confidence: Low Reviewed Nov 08, 2026 Reviewed revision 8775147e1d17 New commits since this review; the score may not cover them
See the full review method →
Maintain this project? Add the FARS badge to your README

The badge shows the current review score and verdict and links back to this page. It refreshes about once an hour after a re-review.

FARS

[![FARS review](https://followagents.com/badge/ai-research-skills.svg)](https://followagents.com/en/agents/ai-research-skills)
<a href="https://followagents.com/en/agents/ai-research-skills"><img src="https://followagents.com/badge/ai-research-skills.svg" alt="FARS review"></a>

Recent releases

Latest entries from the project's GitHub Releases. Notes are quoted from upstream as written, not rewritten or tested by FollowAgents.

  1. v1.7.18 v1.7.18 — research-workspace 0.5.20 alignment
    Release notes excerpt
    Align research-workspace with accepted source plugin 0.5.20; catalog version 1.7.18. Refresh immutable source identity, marketplace cache metadata, bilingual current references and active verification guards. Preserve the prior 0.5.18 review, older history and all other plugin/source identities. Verification:358catalog tests passed,2skipped; source version/layout guards4passed; exact25-path changed-source check and strict live17-skill/5-source metadata health pass. These bounded checks do not establish research quality or universal agent …
    View on GitHub ↗
  2. v1.7.17 v1.7.17 — Research-workspace 0.5.18 alignment
    Release notes excerpt
    Research-workspace version alignment Align the catalog with the accepted research-workspace manifest 0.5.18 at research-hub commit c1ab382e1bc4bfdb1d5df6a230d1ab4509a50d2a, replacing the stale 0.5.14 metadata. Update the current accepted-source record, active guards and bilingual references while preserving prior evidence and every other plugin/source record. …
    View on GitHub ↗
  3. v1.7.16 v1.7.16 — Academic writing 1.3.24 alignment
    Release notes excerpt
    Academic writing source alignment Align academic-writing-skills with accepted release 1.3.24 at 586a5aa5cce3c32eb69629bfc0a631e95867c769 (upstream PR #46). The existing walkthrough and cold-reader procedure now explicitly check the connection from a result to its explanation, preserving clear short sentences and unsupported-mechanism boundaries. Earlier research-identity and concern-closure improvements remain included. …
    View on GitHub ↗

FAQ

Does it work with Chinese-language literature?
Yes. Skills are language-agnostic—run them in Chinese and the output is Chinese; NotebookLM, Claude, and Gemini all handle Traditional/Simplified Chinese well. Stage 1–2 and Stage 2 deliverables also ship with .docx generators.
Is anything paid? What accounts do I need?
The catalog is MIT-licensed and free, but skills run inside your own LLM sessions/keys; all plugins are free. Zotero, Obsidian, and NotebookLM are all optional in research-hub—you can use any subset or none at all.
Are Zotero operations safe?
zotero-library-curator is read-only and emits a preview report; zotero-skills can write (merge duplicates, delete items, rebind collections). The README requires exporting a Zotero RDF backup before letting any AI modify your library—the skills will not back it up for you.
Can I use it without Claude Code?
Yes. The core is the portable SKILL.md: Codex CLI/Cursor/Windsurf load it into a skills/rules directory or inline it into the prompt; Gemini CLI and generic API clients inline it as context; Hermes 0.13.0 is verified. But auto-triggering and marketplace install are fully available only in Claude Code.
How long do typical tasks take?
Maintainer-observed ranges: comparing 5 papers takes 1–3 minutes; a 3-gate gap decision 5–15 minutes; a single-paragraph prose audit under 1 minute; a reviewer response (6 comments) 3–8 minutes. Actual times depend on your LLM provider, network, and prompt phrasing.
Is AI Research Skills open source?
Yes. The source code is on GitHub (WenyuChiou/ai-research-skills) under the MIT licence.
Is AI Research Skills still maintained?
Yes. The latest commit on GitHub is from 2026-11-16, within the last 30 days. The latest release is v1.7.18 (2026-10-29).
Who makes AI Research Skills?
AI Research Skills is published on GitHub by the account WenyuChiou. FollowAgents has not verified which person or company is behind that account.
Was this page helpful?
View on GitHub ↗ Install ↓

Compare agents like this one

The same FARS review applied across the shortlist this agent qualifies for.

Related concepts