Menu

Not in the top 1,000 by GitHub stars

wanshuiyin/Anti-Autoresearch

Don't trust an autoresearch paper at face value. Reviewer-side integrity forensics (self-consistency + fabrication), deterministic verdict. 61 signals: 46 integrity hack-patterns (families A–H, verdict-bearing) + 13 zero-weight AI writing-style impressions (AIS) + 2 advisory. Not an opaque AI-text classifier. The dual of ARIS.

WritingResearch

History

Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.

View weekly changes

skills/adversarial-case-builder/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

45135aa38cbc → 014a1e7cd63f

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 45135aa38cbc

No earlier version was observed for comparison.

Compare on GitHub

skills/ai-style-impressions/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

45135aa38cbc → 014a1e7cd63f

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 45135aa38cbc

No earlier version was observed for comparison.

Compare on GitHub

skills/baseline-comparison-audit/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

45135aa38cbc → 014a1e7cd63f

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 45135aa38cbc

No earlier version was observed for comparison.

Compare on GitHub

skills/citation-forensics/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

45135aa38cbc → 014a1e7cd63f

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 45135aa38cbc

No earlier version was observed for comparison.

Compare on GitHub

skills/consistency-audit/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

45135aa38cbc → 014a1e7cd63f

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 45135aa38cbc

No earlier version was observed for comparison.

Compare on GitHub

skills/eval-design-forensics/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

45135aa38cbc → 014a1e7cd63f

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 45135aa38cbc

No earlier version was observed for comparison.

Compare on GitHub

skills/evidence-ledger/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

45135aa38cbc → 014a1e7cd63f

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 45135aa38cbc

No earlier version was observed for comparison.

Compare on GitHub

skills/experiment-forensics/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

45135aa38cbc → 014a1e7cd63f

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 45135aa38cbc

No earlier version was observed for comparison.

Compare on GitHub

skills/novelty-duplication-advisory/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

45135aa38cbc → 014a1e7cd63f

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 45135aa38cbc

No earlier version was observed for comparison.

Compare on GitHub

skills/presentation-signals/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

45135aa38cbc → 014a1e7cd63f

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 45135aa38cbc

No earlier version was observed for comparison.

Compare on GitHub

skills/proof-derivation-forensics/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

45135aa38cbc → 014a1e7cd63f

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 45135aa38cbc

No earlier version was observed for comparison.

Compare on GitHub

workflows/anti-autoresearch/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

45135aa38cbc → 014a1e7cd63f

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 45135aa38cbc

No earlier version was observed for comparison.

Compare on GitHub

View on GitHubAdd to my skills

Verified commit 014a1e7cd63f1656603daaeeadb856aed1df6869

179GitHub stars
7stars in 7 days
12skills in this repository
5commits in 30 days

Heat

4.7
  • Star growth, last 7 daysNot available
  • Star growth, last 24 hoursNot available
  • Rank movement, last 7 daysNot available
  • Commits, last 30 daysNot available
  • Accounts asking about it through MCP, last 7 daysNot available
  • Hot votes this weekNot available

Heat never changes the order by stars, and sponsorship never touches it. How heat works

Stars over time

Oct 25, 2026Nov 5, 2026179

30-day change not yet available: not enough history.

Individual skills in this repository

How it works

adversarial-case-builder

When your agent uses it

Synthesize the single strongest EVIDENCE-BOUND reviewer case to reject a paper, built ONLY from the evidence ledger (claims.json) + the other auditors' confirmed findings — never free-floating LLM critique. Two fresh cross-model codex threads: an attack writes the ~200-word rejection paragraph (every accusation tagged to an existing claim_id/finding_id), a defense decomposes it and rules each point against the anchored evidence. MEMO-ONLY: emits adversarial-case-builder.memo.md (fed to the adjudicator via --memo) and carries NO verdict weight — tools/adjudicate_findings.py lists it in ZERO_WEIGHT_SKILLS and caps it at info. Honest-null allowed (the paper may survive). Run LAST. Detect-only. Adapted from ARIS kill-argument. Triggers: "adversarial case", "strongest objection", "rejection memo", "kill argument", "最强拒稿点".

What it loads into context

Frontmatter

228 tokens

Always available to the agent

SKILL.md

11,691 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

  1. Why this exists
  2. Core principle
  3. How this differs from the other auditors (route correctly)
  4. How this differs from ARIS `kill-argument` (the parent)
  5. Constants & Reviewer Calling Convention
  6. Step 0 — Preconditions: locate the ledger + the upstream findings, read the level
  7. Step 1 — Attack (Thread 1, fresh codex, evidence-bound ~200 words)
  8. Step 2 — Defense (Thread 2, fresh codex, decompose + classify)
  9. Step 3 — Validate + anchor, then render the memo (the anti-slop gate)
  10. Step 4 — Trace (forensic; never silently dropped)
  11. Step 5 — Hand off, or adjudicate standalone
  12. Output contract
  13. Key rules
  14. When NOT to use this skill
  15. Review tracing

Scripts executed separately

  • ../../tools/adjudicate_findings.py

Files

  • SKILL.mdcore45,832 bytes
  • SKILL.mdcore45,832 bytes
  • SKILL.mdcore45,832 bytes
  • SKILL.mdcore45,832 bytes
  • SKILL.mdcore45,832 bytes
  • SKILL.mdcore45,832 bytes
  • SKILL.mdcore45,832 bytes
  • ../../tools/adjudicate_findings.pyruns46,536 bytes
Read SKILL.md

# Adversarial Case Builder — the strongest *evidence-bound* objection

Build the single strongest evidence-bound case to reject **$ARGUMENTS**, then defend
it point-by-point. Emit `adversarial-case-builder.memo.md`. Run **LAST**, AFTER
`/evidence-ledger` (so `claims.json` exists) and AFTER the auditor skills (so the
merged `*.findings.json` exist).

> 🔒 **Do not wrap this skill in `/loop`, `/schedule`, or `CronCreate`.** It runs
> LAST and synthesizes the ledger + the other auditors' findings into one memo. Even
> though it is **memo-only** (the adjudicator caps it at `info`, so it adds no verdict
> weight), the no-new-signal cadence rule still applies: its output changes only when
> the **ledger / the findings / the paper** change, never with the wall clock.
> Schedule the *work that precedes it* — ledger + auditors done → run this **once**.
> (Mirrors ARIS's external-cadence doctrine.)
View the complete source on GitHub

Skills included

  • adversarial-case-builder
  • ai-style-impressions
  • baseline-comparison-audit
  • citation-forensics
  • consistency-audit
  • eval-design-forensics
  • evidence-ledger
  • experiment-forensics
  • novelty-duplication-advisory
  • presentation-signals
  • proof-derivation-forensics
  • anti-autoresearch

Repository facts

License
MIT
Created
Jun 26, 2026
Last pushed
Nov 6, 2026
Forks
9
Open issues
0
Project activity
Active

Before you install

Network or broad shell

Automated signals, not a guarantee

Checked Nov 7, 2026

Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.

Show 206 notes
  • skills/adversarial-case-builder/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/adversarial-case-builder/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/adversarial-case-builder/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/adversarial-case-builder/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/adversarial-case-builder/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/adversarial-case-builder/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/ai-style-impressions/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/ai-style-impressions/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/ai-style-impressions/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/ai-style-impressions/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/ai-style-impressions/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/ai-style-impressions/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/ai-style-impressions/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/ai-style-impressions/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/ai-style-impressions/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/ai-style-impressions/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/ai-style-impressions/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/ai-style-impressions/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/baseline-comparison-audit/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/baseline-comparison-audit/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
Show all 76 signals by file

skills/adversarial-case-builder/SKILL.md

  • Broad permissionsline 1
    Bash(*), Read, Write, mcp__codex__codex

skills/ai-style-impressions/SKILL.md

  • Broad permissionsline 1
    Bash(*), Read, Write, mcp__codex__codex

skills/baseline-comparison-audit/SKILL.md

  • Broad permissionsline 1
    Bash(*), Read, Write, Grep, Glob, WebSearch, WebFetch, mcp__codex__codex
  • Network accessline 320
    "profile_version": "0.1",   "profile_hit": true,   "leaderboard_source": "https://paperswithcode.com/sota/arithmetic-reasoning-on-gsm8k  (read <UTC date>)",
  • Network accessline 321
    "profile_hit": true,   "leaderboard_source": "https://paperswithcode.com/sota/arithmetic-reasoning-on-gsm8k  (read <UTC date>)",   "expected_baselines": [
  • Network accessline 322
    "leaderboard_source": "https://paperswithcode.com/sota/arithmetic-reasoning-on-gsm8k  (read <UTC date>)",

skills/citation-forensics/SKILL.md

  • Broad permissionsline 1
    Bash(*), Read, Write, Grep, Glob, WebSearch, WebFetch, mcp__codex__codex, mcp__mcp-dblp__search, mcp__mcp-dblp__fuzzy_title_search, mcp__mcp-dblp__get_venue_info
  • Network accessline 308
    Journal / Repository). - **arXiv / DOI resolve + content** — `WebFetch` the arXiv abstract page built from   the bib `eprint`/`arxiv` id (`https://arxiv.org/abs/<id>`) and capture the returned
  • Network accessline 309
    - **arXiv / DOI resolve + content** — `WebFetch` the arXiv abstract page built from   the bib `eprint`/`arxiv` id (`https://arxiv.org/abs/<id>`) and capture the returned   title + authors **and the abstract text** (the abstract is what lets
  • Network accessline 310
    the bib `eprint`/`arxiv` id (`https://arxiv.org/abs/<id>`) and capture the returned
  • Network accessline 311
    title + authors **and the abstract text** (the abstract is what lets the reviewer   judge *context*); `WebFetch https://doi.org/<doi>` to confirm the DOI resolves to   this work. Use `WebSearch` as a fallback for very recent papers (< ~2 we
  • Network accessline 312
    judge *context*); `WebFetch https://doi.org/<doi>` to confirm the DOI resolves to
  • Network accessline 322
    "smith2024bar": {     "dblp_fuzzy_top": [{"title": "...", "authors": ["..."], "venue": "...",                         "year": 2024, "doi": "...", "url": "https://dblp.org/..."}],
  • Network accessline 323
    "dblp_fuzzy_top": [{"title": "...", "authors": ["..."], "venue": "...",                         "year": 2024, "doi": "...", "url": "https://dblp.org/..."}],     "venue_info": {"venue": "...", "type": "Conference or Workshop"},
  • Network accessline 324
    "year": 2024, "doi": "...", "url": "https://dblp.org/..."}],
  • Network accessline 493
    "finding_id": "F001", "skill": "citation-forensics", "pattern_id": "HP-CITE-CONTEXT",   "title": "Self-refinement work cited for the opposite of what it shows",   "description": "The sentence cites \\cite{madaan2023selfrefine} to support th
  • Network accessline 494
    "title": "Self-refinement work cited for the opposite of what it shows",   "description": "The sentence cites \\cite{madaan2023selfrefine} to support that self-feedback yields correlated errors. The cited paper (Self-Refine, NeurIPS 2023; D
  • Network accessline 495
    "description": "The sentence cites \\cite{madaan2023selfrefine} to support that self-feedback yields correlated errors. The cited paper (Self-Refine, NeurIPS 2023; DBLP https://dblp.org/... confirms title/venue/authors; abstract in resoluti

skills/consistency-audit/SKILL.md

  • Broad permissionsline 1
    Bash(*), Read, Write, mcp__codex__codex

skills/eval-design-forensics/SKILL.md

  • Broad permissionsline 1
    Bash(*), Read, Write, Grep, Glob, WebSearch, WebFetch, mcp__codex__codex

skills/evidence-ledger/SKILL.md

  • Broad permissionsline 1
    Bash(*), Read, Write, Grep, Glob, mcp__codex__codex
  • Network accessline 118
    PAPER_ID="${ARG//./_}"     PAPER_DIR="$(pwd)/.aa_work/$PAPER_ID"; mkdir -p "$PAPER_DIR"     ( curl -fsSL "https://arxiv.org/e-print/$ARG" -o "$PAPER_DIR/src.tar" \
  • Network accessline 119
    PAPER_DIR="$(pwd)/.aa_work/$PAPER_ID"; mkdir -p "$PAPER_DIR"     ( curl -fsSL "https://arxiv.org/e-print/$ARG" -o "$PAPER_DIR/src.tar" \         && tar -xf "$PAPER_DIR/src.tar" -C "$PAPER_DIR" 2>/dev/null ) \
  • Network accessline 120
    ( curl -fsSL "https://arxiv.org/e-print/$ARG" -o "$PAPER_DIR/src.tar" \
  • Network accessline 121
    && tar -xf "$PAPER_DIR/src.tar" -C "$PAPER_DIR" 2>/dev/null ) \       || curl -fsSL "https://arxiv.org/pdf/$ARG.pdf" -o "$PAPER_DIR/paper.pdf"   # source preferred (better spans) else
  • Network accessline 122
    || curl -fsSL "https://arxiv.org/pdf/$ARG.pdf" -o "$PAPER_DIR/paper.pdf"   # source preferred (better spans)

skills/experiment-forensics/SKILL.md

  • Broad permissionsline 1
    Bash(*), Read, Write, Grep, Glob, mcp__codex__codex
  • Dangerous shell operationline 34
    1. **Fake ground truth** — the eval "reference/target" is *derived from model

skills/novelty-duplication-advisory/SKILL.md

  • Broad permissionsline 1
    Bash(*), Read, Write, WebSearch, WebFetch, mcp__codex__codex, mcp__mcp-dblp__fuzzy_title_search, mcp__mcp-dblp__search
  • Dangerous shell operationline 740
    rm = json.load(open(runmeta, encoding="utf-8"))

skills/presentation-signals/SKILL.md

  • Broad permissionsline 1
    Bash(*), Read, Write, mcp__codex__codex

skills/proof-derivation-forensics/SKILL.md

  • Broad permissionsline 1
    Bash(*), Read, Write, mcp__codex__codex

tools/adjudicate_findings.py

  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3

tools/attach_refutation.py

  • Executable scriptline 1
    #!/usr/bin/env python3
  • Access to secretsline 70
    [Credential-related source omitted]
  • Access to secretsline 71
    [Credential-related source omitted]

tools/build_claim_ledger.py

  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3

tools/build_manifest.py

  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3

tools/check_ai_style.py

  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3

tools/check_numeric_consistency.py

  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3

tools/check_presentation.py

  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3

tools/check_stat_consistency.py

  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3
  • Dangerous shell operationline 89
    r"|\b(?:test|eval(?:uation)?|held[- ]?out|dev|validation)\s*set\s*of\s*(\d{1,7})\b"
  • Dangerous shell operationline 89
    r"|\b(?:test|eval(?:uation)?|held[- ]?out|dev|validation)\s*set\s*of\s*(\d{1,7})\b"

tools/select_primary_pdf.py

  • Executable scriptline 1
    #!/usr/bin/env python3
  • Executable scriptline 1
    #!/usr/bin/env python3

workflows/anti-autoresearch/SKILL.md

  • Broad permissionsline 1
    Bash(*), Read, Write, Grep, Glob, Skill, WebSearch, WebFetch, mcp__codex__codex, mcp__mcp-dblp__search, mcp__mcp-dblp__fuzzy_title_search, mcp__mcp-dblp__get_venue_info
  • Network accessline 278
    PAPER_DIR="$(pwd)/aar-$ID"; mkdir -p "$PAPER_DIR"   # LaTeX source first (best spans → L1). e-print may be a tarball OR a single gzipped .tex.   if curl -fsSL "https://arxiv.org/e-print/$ID" -o "$PAPER_DIR/src.tgz"; then
  • Network accessline 279
    # LaTeX source first (best spans → L1). e-print may be a tarball OR a single gzipped .tex.   if curl -fsSL "https://arxiv.org/e-print/$ID" -o "$PAPER_DIR/src.tgz"; then     tar -xzf "$PAPER_DIR/src.tgz" -C "$PAPER_DIR" 2>/dev/null \
  • Network accessline 280
    if curl -fsSL "https://arxiv.org/e-print/$ID" -o "$PAPER_DIR/src.tgz"; then
  • Network accessline 283
    fi   if ! ls "$PAPER_DIR"/*.tex >/dev/null 2>&1; then     # fallback: PDF → text (L0)     curl -fsSL "https://arxiv.org/pdf/$ID.pdf" -o "$PAPER_DIR/paper.pdf" \
  • Network accessline 284
    if ! ls "$PAPER_DIR"/*.tex >/dev/null 2>&1; then     # fallback: PDF → text (L0)     curl -fsSL "https://arxiv.org/pdf/$ID.pdf" -o "$PAPER_DIR/paper.pdf" \       && pdftotext -layout "$PAPER_DIR/paper.pdf" "$PAPER_DIR/paper.txt"
  • Network accessline 285
    curl -fsSL "https://arxiv.org/pdf/$ID.pdf" -o "$PAPER_DIR/paper.pdf" \
  • Dangerous shell operationline 479
    from a PDF. Its full line-by-line code audit only fires at **L2**; the eval `file:line`

How the signals fit together

Facts from a static read of the files at the pinned commit. They describe what the code does, not why.

Some files could not be read; the facts below still hold.

Signal confidence: 76 facts

and 26 more

Works with

Agent compatibility, from the files in this repository
AgentStatusEvidenceInstall to
Claude CodeNativeFrontmatter field argument-hint in skills/adversarial-case-builder/SKILL.md.claude/skills/<skill-name>/
CodexCompatibleStandard SKILL.md skills/adversarial-case-builder/SKILL.md.agents/skills/<skill-name>/
CursorCompatibleStandard SKILL.md skills/adversarial-case-builder/SKILL.md.cursor/skills/<skill-name>/
Gemini CLICompatibleStandard SKILL.md skills/adversarial-case-builder/SKILL.md.gemini/skills/<skill-name>/
GitHub CopilotCompatibleStandard SKILL.md skills/adversarial-case-builder/SKILL.md.github/skills/<skill-name>/
OpenCodeCompatibleStandard SKILL.md skills/adversarial-case-builder/SKILL.md.opencode/skills/<skill-name>/
  • Native The repository ships a manifest, folder or frontmatter field for this agent.
  • Compatible A standard SKILL.md that this agent can read.
  • Unknown No evidence either way.

Derived automatically from paths, manifests and frontmatter. Not tested on each agent.

You might also use

Marazii/research-co-pilot

Comprehensive research co-pilot for Claude Code and claude.ai — 14 skills covering literature review, methodology, ethics review, data analysis, qualitative coding, brainstorming,…

15GitHub stars

K-Dense-AI/scientific-agent-skills

Turn any AI agent into an AI Scientist. The #1 Agent Skills library for science, used by 250,000+ scientists worldwide. 177 ready-to-use validated skills plus 100+ scientific data…

56.4kGitHub stars

wanshuiyin/Auto-claude-code-research-in-sleep

ARIS ⚔️ (Auto-Research-In-Sleep) — Lightweight Markdown-only skills for autonomous ML research: cross-model review loops, idea discovery, and experiment automation. No framework, …

18.8kGitHub stars

Orchestra-Research/AI-Research-SKILLs

Comprehensive open-source library of AI research and engineering skills for any AI model. Package the skills and your claude code/codex/gemini agent will be an AI research agent w…

14.3kGitHub stars

Imbad0202/academic-research-skills-codex

Codex-native Academic Research Skills suite for human-in-the-loop academic research workflows

14.1kGitHub stars

uditgoenka/autoresearch

Claude Autoresearch Skill — Autonomous goal-directed iteration for Claude Code. Inspired by Karpathy's autoresearch. Modify → Verify → Keep/Discard → Repeat forever.

6.8kGitHub stars