Deep Research — Universal Academic Research Agent Team
Universal deep research agent team. 13-agent pipeline for rigorous academic research on any topic. 8 modes: full research, quick brief…
You say
- Context cost
- 188.5k tokensestimated from the bundle, loaded when it triggers
- Bundle
- 53 files · 754.1 kBtext throughout, nothing executable
- Licence
- Source-availablefree to use
- Last change
- —v2.12.1
- Servers it uses
- Noneruns standalone
What it does
Universal deep research agent team. 13-agent pipeline for rigorous academic research on any topic. 8 modes: full research, quick brief, paper review, lit-review, fact-check, three-way literature scan, Socratic guided research dialogue, and systematic review with optional meta-analysis. Covers research question formulation, Socratic mentoring, methodology design, systematic literature search, source verification, cross-source synthesis, risk of bias assessment, meta-analysis, APA 7.0 report compilation, editorial review, devil's advocate challenges, ethics review, and post-research literature…
Installed, it changes the agent in these ways.
What this skill changes about the agent is not written down here yet. The listing was collected from its source, and the description is in its own SKILL.md.
And these are the jobs people reach for it for.
Workflow
Runs a procedure end to end.
Filed under
The skill itself
This is the whole product. A skill is instructions the model reads, so there is nothing behind the listing you cannot see first — the front matter loads with every session, and the body below it loads when the skill triggers.
SKILL.md37.0 kB · 601 lines
---
name: deep-research
description: "Universal deep research agent team. 13-agent pipeline for rigorous academic research on any topic. 8 modes: full research, quick brief, paper review, lit-review, fact-check, three-way literature scan, Socratic guided research dialogue, and systematic review with optional meta-analysis. Covers research question formulation, Socratic mentoring, methodology design, systematic literature search, source verification, cross-source synthesis, risk of bias assessment, meta-analysis, APA 7.0 report compilation, editorial review, devil's advocate challenges, ethics review, and post-research literature monitoring. Triggers on: research, deep research, literature review, systematic review, meta-analysis, PRISMA, evidence synthesis, fact-check, WHY HOW WHAT papers, 3W literature scan, guide my research, help me think through, 研究, 深度研究, 文獻回顧, 文獻探討, 系統性回顧, 後設分析, 事實查核, 三段式文獻掃描, 引導我的研究, 幫我釐清, 幫我想想, 我不確定要研究什麼, 研究方向, 研究主題, 심층 연구, 문헌 조사, 체계적 문헌고찰, 메타분석, 사실 확인, 연구 방향을 잡아줘, 연구 주제 정하는 것을 도와줘."
metadata:
version: "2.12.1"
last_updated: "2026-08-15"
status: active
data_access_level: raw
task_type: open-ended
related_skills:
- academic-paper
- academic-pipeline
---
15# Deep Research — Universal Academic Research Agent Team
16
17Universal deep research tool — a domain-agnostic 13-agent team for rigorous academic research on any topic.
18
19**v2.4** adds writing quality improvements to the report compiler:
20- **Style Profile consumption** (optional) — If a Style Profile is available from academic-paper intake, the report compiler applies it as a soft guide for the Executive Summary and Synthesis sections. Discipline conventions and report objectivity take priority.
21- **Writing Quality Check** — The report compiler runs a writing quality checklist before finalizing: flags AI-typical overused terms, checks sentence/paragraph length variation, removes throat-clearing openers. See academic-paper/references/writing_quality_check.md.
22
23> **Routing discipline (v3.9.2):** see .claude/CLAUDE.md "Routing Discipline (v3.9.2)" + shared/references/intent_clarification_protocol.md for cross-skill routing rules. This skill assumes routing has already settled — ambiguous cross-phase materials should have been clarified upstream.
24
25## Quick Start
26
27**Minimal command:**
28```
29Research the impact of AI on higher education quality assurance
30```
31
32**Socratic mode:**
33```
34Guide my research on the impact of declining birth rates on private universities
35引導我的研究：少子化對私立大學的影響
36幫我釐清我的研究方向，我對高教品保有興趣但還不太確定
37```
38
39**Execution:**
401. Scoping — Research question + methodology blueprint
412. Investigation — Systematic literature search + source verification
423. Analysis — Cross-source synthesis + bias check
434. Composition — Full APA 7.0 report
445. Review — Editorial + ethics + vulnerability scan
456. Revision — Final polished report
46
47---
48
49## Trigger Conditions
50
51### Trigger Keywords
52
53**English**: research, deep research, literature review, systematic review, meta-analysis, PRISMA, evidence synthesis, fact-check, methodology, APA report, academic analysis, policy analysis, WHY HOW WHAT papers, 3W literature scan, guide my research, help me think through, monitor this topic, set up alerts
54
55**繁體中文**: 研究, 深度研究, 文獻回顧, 文獻探討, 系統性回顧, 後設分析, 證據綜整, 事實查核, 三段式文獻掃描, WHY HOW WHAT 論文比較, 研究方法, 學術分析, 政策分析, 引導我的研究, 幫我釐清, 監測這個主題, 設定追蹤
56
57**한국어**: 심층 연구, 문헌 조사, 문헌 고찰, 체계적 문헌고찰, 메타분석, 근거 종합, 사실 확인, 팩트체크, 연구 방법 설계, 학술 분석, 연구 방향을 잡아줘, 연구 주제 정하는 것을 도와줘, 무엇을 연구할지 모르겠어, 이 주제 계속 모니터링해줘
58
59### Socratic Mode Activation
60
61Activate socratic mode when the user's **intent** matches any of the following patterns, **regardless of language**. Detect meaning, not exact keywords.
62
63**Intent signals** (any one is sufficient):
641. User has no clear research question and wants guided thinking
652. User asks to be "led", "guided", or "mentored" through research
663. User expresses uncertainty about what to research or where to start
674. User wants to brainstorm, explore, or clarify a research direction
685. User describes a vague interest without a specific, answerable question
69
70**Default rule**: When intent is ambiguous between socratic and full, **prefer socratic** — it is safer to guide first than to produce an unwanted report. The user can always switch to full later.
71
72**Example triggers** (illustrative, not exhaustive):
73"guide my research", "help me think through", 「引導我的研究」「幫我釐清」, or equivalent in any language
74
75### Does NOT Trigger
76
77| Scenario | Use Instead |
78|----------|-------------|
79| Writing a paper (not researching) | academic-paper |
80| Reviewing a paper (structured review) | academic-paper-reviewer |
81| Full research-to-paper pipeline | academic-pipeline |
82
83### Quick Mode Selection Guide
84
85| Your Situation 你的狀況 | Recommended Mode | Spectrum |
86|----------------|-----------------|----------|
87| Vague idea, need guidance / 有模糊想法，需要引導 | socratic | originality |
88| Clear RQ, need comprehensive research / 有明確 RQ，需要完整研究 | full | balanced |
89| Need a quick brief (30 min) / 需要快速摘要 | quick | fidelity |
90| Have a paper to evaluate before citing / 有論文需要評估 | review | balanced |
91| Need literature review for a topic / 需要文獻回顧 | lit-review | fidelity |
92| Need a fast paper-comparison scan / 需要快速比較多篇論文 | three-way-scan | fidelity |
93| Need to verify specific claims / 需要查核特定事實 | fact-check | fidelity |
94| Need systematic review / meta-analysis / 系統性回顧或後設分析 | systematic-review | fidelity |
95
96**Spectrum** (v3.2): *fidelity* = template-heavy, predictable output; *balanced* = default; *originality* = exploratory, template-light. See shared/mode_spectrum.md for the full cross-skill spectrum table.
97
98Not sure? Start with socratic — it will help you figure out what you need.
99不確定？先用 socratic 模式——它會幫你釐清你需要什麼。
100
101---
102
103## Agent Team (13 Agents)
104
105| # | Agent | Role | Phase |
106|---|-------|------|-------|
107| 1 | research_question_agent | Transforms vague topics into precise, FINER-scored research questions with scope boundaries | Phase 1, Socratic Layer 1 |
108| 2 | research_architect_agent | Designs methodology blueprint: paradigm, method, data strategy, analytical framework, validity criteria | Phase 1 |
109| 3 | bibliography_agent | Systematic literature search, source screening, annotated bibliography in APA 7.0 | Phase 2 |
110| 4 | source_verification_agent | Fact-checking, source grading (evidence hierarchy), predatory journal detection, conflict-of-interest flagging | Phase 2 |
111| 5 | synthesis_agent | Cross-source integration, contradiction resolution, thematic synthesis, gap analysis | Phase 3 |
112| 6 | report_compiler_agent | Drafts complete APA 7.0 report (Title -> Abstract -> Intro -> Method -> Findings -> Discussion -> References) | Phase 4, 6 |
113| 7 | editor_in_chief_agent | Q1 journal editorial review: originality, rigor, evidence sufficiency, verdict (Accept/Revise/Reject) | Phase 5 |
114| 8 | devils_advocate_agent | Challenges assumptions, tests for logical fallacies, finds alternative explanations, confirmation bias checks | Phase 1, 3, 5, Socratic Layer 2, 4 |
115| 9 | ethics_review_agent | AI-assisted research ethics, attribution integrity, dual-use screening, fair representation | Phase 5 |
116| 10 | socratic_mentor_agent | Q1 journal editor persona; guides research thinking through Socratic questioning across 5 layers | Socratic Mode (Layer 1-5) |
117| 11 | risk_of_bias_agent | Assesses risk of bias using RoB 2 (RCTs) and ROBINS-I (non-randomized); traffic-light visualization | Systematic Review (Phase 2) |
118| 12 | meta_analysis_agent | Designs and executes meta-analysis or narrative synthesis; effect sizes, heterogeneity, GRADE | Systematic Review (Phase 3) |
119| 13 | monitoring_agent | Post-research literature monitoring: digests, retraction alerts, contradictory findings detection | Optional (post-pipeline) |
120
121---
122
123## Mode Selection Guide
124
125See references/mode_selection_guide.md for the detailed guide.
126
127```
128User Input
129 |
130 +-- Already have a clear research question?
131 | +-- Yes --> Need PRISMA-compliant systematic review / meta-analysis?
132 | | +-- Yes --> systematic-review mode
133 | | +-- No --> Need a full report?
134 | | +-- Yes --> full mode
135 | | +-- No --> Only need literature?
136 | | +-- Yes --> Need rapid paper comparison?
137 | | +-- Yes --> three-way-scan mode
138 | | +-- No --> lit-review mode
139 | | +-- No --> quick mode
140 | +-- No --> Want to be guided through thinking?
141 | +-- Yes --> socratic mode
142 | +-- No --> full mode (Phase 1 will be interactive)
143 |
144 +-- Already have text to review? --> review mode
145 +-- Only need fact-checking? --> fact-check mode
146```
147
148---
149
150## Orchestration Workflow (6 Phases)
151
152```
153User: "Research [topic]"
154 |
155=== Phase 1: SCOPING (Interactive) ===
156 |
157 |-> [research_question_agent] -> RQ Brief
158 | - FINER criteria scoring (Feasible, Interesting, Novel, Ethical, Relevant)
159 | - Scope boundaries (in-scope / out-of-scope)
160 | - 2-3 sub-questions
161 |
162 |-> [research_architect_agent] -> Methodology Blueprint
163 | - Research paradigm (positivist / interpretivist / pragmatist)
164 | - Method selection (qualitative / quantitative / mixed)
165 | - Data strategy (primary / secondary / both)
166 | - Analytical framework
167 | - Validity & reliability criteria
168 |
169 +-> [devils_advocate_agent] -- CHECKPOINT 1
170 - RQ clarity and answerable?
171 - Method appropriate for question?
172 - Scope too broad or too narrow?
173 - Verdict: PASS / REVISE (with specific feedback)
174 |
175 ** User confirmation before Phase 2 **
176 |
177=== Phase 2: INVESTIGATION ===
178 |
179 |-> [bibliography_agent] -> Source Corpus + Annotated Bibliography
180 | - Systematic search strategy (databases, keywords, Boolean)
181 | - Inclusion/exclusion criteria
182 | - PRISMA-style flow (if applicable)
183 | - Annotated bibliography (APA 7.0)
184 |
185 +-> [source_verification_agent] -> Verified & Graded Sources
186 - Evidence hierarchy grading (Level I-VII)
187 - Predatory journal screening
188 - Conflict-of-interest flagging
189 - Currency assessment (publication date relevance)
190 - Source quality matrix
191 |
192=== Phase 3: ANALYSIS ===
193 |
194 |-> [synthesis_agent] -> Synthesis Narrative + Gap Analysis
195 | - Thematic synthesis across sources
196 | - Contradiction identification & resolution
197 | - Evidence convergence/divergence mapping
198 | - Knowledge gap analysis
199 | - Theoretical framework integration
200 |
201 +-> [devils_advocate_agent] -- CHECKPOINT 2
202 - Cherry-picking check
203 - Confirmation bias detection
204 - Logic chain validation
205 - Alternative explanations explored?
206 - Verdict: PASS / REVISE
207 |
208=== Phase 4: COMPOSITION ===
209 |
210 +-> [report_compiler_agent] -> Full APA 7.0 Draft
211 - Title Page
212 - Abstract (150-250 words)
213 - Introduction (context, problem, purpose, RQ)
214 - Literature Review / Theoretical Framework
215 - Methodology
216 - Findings / Results
217 - Discussion (interpretation, implications, limitations)
218 - Conclusion & Recommendations
219 - References (APA 7.0)
220 - Appendices (if applicable)
221 |
222=== Phase 5: REVIEW (Parallel) ===
223 |
224 |-> [editor_in_chief_agent] -> Editorial Verdict + Line Feedback
225 | - Originality assessment
226 | - Methodological rigor
227 | - Evidence sufficiency
228 | - Argument coherence
229 | - Writing quality (clarity, conciseness, flow)
230 | - Verdict: ACCEPT / MINOR REVISION / MAJOR REVISION / REJECT
231 |
232 |-> [ethics_review_agent] -> Research-Integrity Review + Human-Subjects Administrative Status
233 | - AI disclosure compliance
234 | - Attribution integrity
235 | - Dual-use screening
236 | - Fair representation check
237 | - Integrity verdict only: CLEARED / CONDITIONAL / BLOCKED
238 | - Human subjects: readiness and authorization reported separately; institutional determination required
239 | - Authority-bound planning: exact requirement IDs + actor/consumer scope only after the #666 replay-validated resolved-context gate
240 | - Candidate rule trace: display only a replay-validated and surface-linted #669 artifact; never use it as a pathway result or workflow input
241 | - Packet structure: consume only a replay-validated #667 manifest; deterministic status never becomes authorization or content adequacy
242 | - Content coverage: consume only a replay-validated #681 LLM-ADVISORY; preserve deterministic status and report efficacy as UNMEASURED
243 |
244 +-> [devils_advocate_agent] -- CHECKPOINT 3
245 - Final vulnerability scan
246 - Strongest counter-argument test
247 - "So what?" significance check
248 - Verdict: PASS / REVISE
249 |
250=== Phase 6: REVISION ===
251 |
252 +-> [report_compiler_agent] -> Final Report
253 - Address editorial feedback
254 - Resolve ethics conditions
255 - Incorporate devil's advocate insights
256 - Max 2 revision loops
257 - Remaining issues -> "Acknowledged Limitations" section
258```
259
260### Checkpoint Rules
261
2621. ⚠️ **IRON RULE**: **Devil's Advocate** has 3 mandatory checkpoints; **Critical-severity** issues block progression
2632. Revision loops capped at **2 iterations**; remaining issues become "acknowledged limitations"
2643. ⚠️ **IRON RULE**: **Ethics Review** stops the user once to confirm a Critical **integrity** concern (fabrication / plagiarism / missing AI disclosure / source misrepresentation / concrete harm-enabling specifics). Overridable with recorded reasoning — it confirms, it does not veto. Subject matter alone never blocks; dual-use is advisory (Responsible Use Statement), not a block.
2654. User confirmation required after Phase 1 before proceeding
266
267---
268
269## Phase-by-phase Invocation Contract (v3.9.2)
270
271ARS pipeline runs in 6 phases. Two invocation modes:
272
273**Mode A — orchestrator-driven (default):** pipeline_orchestrator_agent (in academic-pipeline skill) runs all phases end-to-end with state tracking via Material Passport.
274
275**Mode B — phase-by-phase (cross-session resume):** User invokes one agent per phase across sessions for long-running projects. Common pattern via ARS_PASSPORT_RESET=1 + resume_from_passport=<hash> (see academic-pipeline/references/passport_as_reset_boundary.md).
276
277In Mode B, **single-phase agents (Bucket A per docs/design/2026-05-18-ars-v3.9.2-agent-phase-classification.md) stay strictly within their assigned phase for writes**. Reads from upstream phases are allowed. Multi-phase agents (Bucket B: devils_advocate_agent, report_compiler_agent) do exactly the work specified by the caller's invocation for that phase — no extension to other phases in the same call.
278
279Routing into Mode B requires explicit user signal — /ars-<mode> slash command or [direct-mode] prefix. Ambiguous cross-phase input defaults to clarification per .claude/CLAUDE.md Routing Discipline + shared/references/intent_clarification_protocol.md.
280
281**Enforcement (v3.9.2):** Phase Boundary blocks on Bucket A agents + advisory verifier (scripts/check_pipeline_integrity.py) + a deterministic PreToolUse write-scope guard in hook-enabled runtimes (#134 rescope, PR #294). Multi-phase envelope remains forward-scope (#134 Slices 3-5).
282
283---
284
285## Socratic Mode: Guided Research Dialogue
286
2875-layer dialogue guiding users from vague ideas to concrete research questions. Core principle while non-generation Socratic mode is active: ⚠️ **IRON RULE**: Never give direct answers. The explicit candidate-generation exit below leaves that mode before any candidate is shown.
288
289**Layers**: Clarification -> Assumption Probing -> Evidence/Reasoning -> Viewpoint/Perspective -> Implication/Consequence
290
291**Research-question authorship boundary:** Socratic mode is non-generation by
292default. Non-convergence may produce only a summary of directions the user has
293already expressed plus focused questions or a lit-review suggestion; it never
294produces candidate RQs automatically. If the user explicitly asks the system to
295propose candidates, announce the exit from non-generation Socratic mode and
296emit [SOCRATIC-NON-GENERATION-EXIT: explicit_user_request] on a standalone
297line before any clearly labeled AI-generated candidate. Never switch silently.
298
299> See references/socratic_mode_protocol.md for the full 5-layer dialogue flow, management rules, and auto-end conditions.
300
301### Opt-in Reading Probe (v3.5.1)
302
303Setting ARS_SOCRATIC_READING_PROBE=1 enables a one-time honesty probe during **goal-oriented** Socratic sessions. When the user cites a specific paper, the Mentor asks them to paraphrase one passage. Decline is logged without penalty. Default OFF. See agents/socratic_mentor_agent.md §"Optional Reading Probe Layer".
304
305---
306
307## Systematic Review Mode
308
309PRISMA 2020-compliant systematic review with optional meta-analysis. Follows 5-phase protocol: Protocol Registration -> Systematic Search -> Screening & Selection -> Data Extraction & RoB -> Synthesis & Reporting.
310
311> **v3.4.0 compliance:** systematic-review mode triggers compliance_agent at Stage 2.5 (Methods items) and Stage 4.5 (remaining items + RAISE 8-role matrix). PRISMA-trAIce Mandatory failures block the pipeline. See shared/compliance_checkpoint_protocol.md.
312
313> See references/systematic_review_protocol.md for full PRISMA pipeline, checkpoint rules, and meta-analysis procedures.
314
315---
316
317## Operational Modes
318
319| Mode | Agents Active | Output | Word Count |
320|------|---------------|--------|------------|
321| full (default) | All 9 core (excluding socratic_mentor, RoB, meta-analysis) | Full APA 7.0 report | 3,000-8,000 |
322| quick | RQ + Biblio + Verification + Report | Research brief | 500-1,500 |
323| review | Editor + Devil's Advocate + Ethics | Reviewer report on provided text | N/A |
324| lit-review | Biblio + Verification + Synthesis | Annotated bibliography + synthesis | 1,500-4,000 |
325| three-way-scan | Biblio + Verification (retrieval + WHY/HOW/WHAT extract) | Paper shortlist compared by WHY/HOW/WHAT + cross-paper synthesis | 800-2,000 |
326| fact-check | Source Verification only | Verification report | 300-800 |
327| socratic | Socratic Mentor + RQ + Devil's Advocate | Research Plan Summary (INSIGHT collection) | N/A (iterative) |
328| systematic-review | RQ + Architect + Biblio + Verification + RoB + Meta-Analysis + Synthesis + Report + Editor + Ethics + DA | Full PRISMA 2020 report + forest plot data + GRADE table | 5,000-15,000 |
329
330---
331
332## Three-Way Scan Mode (WHY / HOW / WHAT)
333
334Use three-way-scan when the user needs a disciplined shortlist of papers compared in a stable frame, but does **not** yet need a full literature review report.
335
336- **WHY**: what problem or bottleneck the paper addresses and why it matters
337- **HOW**: what strategy, method, or technical route the paper uses
338- **WHAT**: what the paper found, built, or still leaves unresolved
339
340This mode is intentionally lighter than lit-review. It prioritizes:
341
3421. candidate retrieval
3432. deduplication
3443. compact per-paper extraction
3454. cross-paper synthesis of shared WHY, divergent HOW, and remaining gaps
346
347Recommended per-paper output:
348
349```markdown
350## <paper title>
351Source: <provider> | Year: <year> | Link: <url>
352
353- WHY: ...
354- HOW: ...
355- WHAT: ...
356```
357
358Then add:
359
360- common WHY
361- divergent HOW
362- strongest WHAT
363- unresolved global gap
364
365If the user later wants a broader evidence matrix, thematic synthesis, or PRISMA-like coverage, escalate from three-way-scan to lit-review or systematic-review.
366
367---
368
369## Failure Paths
370
371See references/failure_paths.md for all failure scenarios, trigger conditions, and recovery strategies across all modes.
372
373Key failure path summary:
374
375| Failure Scenario | Trigger Condition | Recovery Strategy |
376|---------|---------|---------|
377| RQ cannot converge | Phase 1 / Layer 1 exceeds multiple rounds while still vague | Full mode may use its candidate workflow; Socratic mode summarizes user-expressed directions or suggests lit-review, with no candidate generation unless the user explicitly exits non-generation mode |
378| Insufficient literature | bibliography_agent finds < 5 sources | Expand search strategy, alternative keywords |
379| Methodology mismatch | RQ type misaligned with method capability | Return to Phase 1, suggest 3 alternative methods |
380| Devil's Advocate CRITICAL | Fatal logical flaw discovered | STOP, explain the issue, require correction |
381| Ethics BLOCKED | Critical integrity issue (not subject matter) | Stop the user once to confirm; list issues + remediation path; overridable with recorded reasoning |
382| Socratic non-convergence | > 10 rounds without convergence | Suggest switching to full mode |
383| User abandons mid-process | Explicitly states they don't want to continue | Save progress, provide re-entry path |
384| Only Chinese-language literature | English search returns empty | Switch to Chinese academic databases |
385
386---
387
388## Literature Monitoring (Optional Post-Pipeline)
389
390Optional post-research monitoring for new publications in the research area.
391
392> See references/literature_monitoring_strategies.md for setup instructions across academic databases.
393
394---
395
396## Handoff Protocol: deep-research → academic-paper
397
398After research is complete, the following materials can be handed off to academic-paper:
399
4001. **Research Question Brief** (from research_question_agent)
4012. **Methodology Blueprint** (from research_architect_agent)
4023. **Annotated Bibliography** (from bibliography_agent)
4034. **Synthesis Report** (from synthesis_agent)
4045. **[If socratic mode] INSIGHT Collection and Research Plan Summary**
4056. **Preregistration handoff** — exactly one builder-produced
406 preregistration-artifact/1.0 sidecar (including an unavailable receipt) and,
407 when status=provided, its explicitly named companion bytes
408
409**Trigger**: User says "now help me write a paper" or "write a paper based on this"
410
411academic-paper's intake_agent will automatically detect available materials and skip redundant steps:
412- Has RQ Brief -> skip topic scoping
413- Has Bibliography -> skip literature search
414- Has Synthesis -> accelerate findings / discussion writing
415- Has preregistration sidecar -> strict-validate it and its named companion,
416 then carry both byte-for-byte; never rebuild it from prose or a template
417
418The non-shell research_architect_agent supplies only the explicit caller
419declaration and companion handle. Before handoff, a shell-capable dispatcher
420must run the named deterministic build-preregistration-artifact subcommand in
421scripts/build_cross_document_consistency_advisory.py, with caller-held RFC3339
422declared_at. Only that builder may create or update the sidecar. A later
423explicit user supply creates a new builder-produced sidecar; omission or silent
424substitution is invalid.
425
426See examples/handoff_to_paper.md for a detailed handoff example.
427
428---
429
430## Full Academic Pipeline
431
432See academic-pipeline/SKILL.md for the complete workflow.
433
434---
435
436## Agent File References
437
438| Agent | Definition File |
439|-------|----------------|
440| research_question_agent | agents/research_question_agent.md |
441| research_architect_agent | agents/research_architect_agent.md |
442| bibliography_agent | agents/bibliography_agent.md |
443| source_verification_agent | agents/source_verification_agent.md |
444| synthesis_agent | agents/synthesis_agent.md |
445| report_compiler_agent | agents/report_compiler_agent.md |
446| editor_in_chief_agent | agents/editor_in_chief_agent.md |
447| devils_advocate_agent | agents/devils_advocate_agent.md |
448| ethics_review_agent | agents/ethics_review_agent.md |
449| socratic_mentor_agent | agents/socratic_mentor_agent.md |
450| risk_of_bias_agent | agents/risk_of_bias_agent.md |
451| meta_analysis_agent | agents/meta_analysis_agent.md |
452| monitoring_agent | agents/monitoring_agent.md |
453
454---
455
456## Reference Files
457
458| Reference | Purpose | Used By |
459|-----------|---------|---------|
460| references/apa7_style_guide.md | APA 7th edition quick reference | report_compiler, editor_in_chief |
461| references/source_quality_hierarchy.md | Evidence pyramid + grading rubric | source_verification, bibliography |
462| references/methodology_patterns.md | Research design templates | research_architect |
463| references/logical_fallacies.md | 30+ fallacies catalog | devils_advocate |
464| references/ethics_checklist.md | AI disclosure, attribution, dual-use | ethics_review |
465| references/interdisciplinary_bridges.md | Cross-discipline connection patterns | synthesis, research_architect |
466| references/socratic_questioning_framework.md | 6 types of Socratic questions + 30+ prompt patterns | socratic_mentor |
467| references/failure_paths.md | 12 failure scenarios with triggers and recovery paths | all agents |
468| references/mode_selection_guide.md | Mode selection flowchart and comparison table | orchestrator |
469| references/irb_decision_tree.md | Portable human-subjects navigation aid; not an authority, universal taxonomy, or pathway determination | ethics_review, research_architect |
470| shared/references/human_subjects_authority_protocol.md | Exact authority selection, replay validation, actor/consumer filtering, and fail-closed resolved-context gate | ethics_review, research_architect |
471| shared/human_subjects_authority_registry.json | Bounded jurisdiction profiles with exact requirement IDs, authority anchors, obligated actors, and consumer scopes | ethics_review, research_architect |
472| shared/contracts/human_subjects/resolved_authority_context.schema.json | Pointer-only resolved-context shape; consumers still require deterministic replay validation | ethics_review, research_architect |
473| shared/references/review_pathway_rule_trace_protocol.md | Candidate-name ownership, exact selected-profile predicate partition, replay, render, surface lint, and non-consumer boundary (#669) | ethics_review, research_architect |
474| shared/contracts/human_subjects/review_pathway_trace_request.schema.json | Closed caller-owned candidate mapping; every selected-profile pathway_trace requirement is accounted for exactly once | dispatching layer |
475| shared/contracts/human_subjects/review_pathway_rule_trace.schema.json | Closed candidate-only predicate trace; replay and surface lint remain mandatory | ethics_review, research_architect |
476| shared/references/submission_packet_manifest_protocol.md | Deterministic packet inventory, authority replay, status, and non-authorization boundary (#667) | ethics_review, research_architect |
477| shared/contracts/human_subjects/submission_packet_manifest.schema.json | Pointer-only deterministic packet-manifest shape; consumers still require exact replay validation | ethics_review, research_architect |
478| shared/references/authority_content_coverage_advisory_protocol.md | Replay-bound authority-profile content observations, evidence-row/1.1 provenance, and noninterference boundary (#681) | ethics_review, research_architect |
479| shared/contracts/human_subjects/content_coverage_advisory.schema.json | Closed LLM-ADVISORY carrier; consumers still require finalizer replay validation | ethics_review, research_architect |
480| shared/contracts/evidence/evidence_row_v1_1.schema.json | Requirement/expectation/artifact-bound bounded excerpt rows for the #681 advisory surface | ethics_review |
481| references/equator_reporting_guidelines.md | EQUATOR reporting guideline mapping | research_architect, report_compiler |
482| references/preregistration_guide.md | Preregistration decision tree + platforms + checklist | research_architect |
483| shared/references/cross_document_consistency_advisory_protocol.md | Exact preregistration sidecar ownership/replay plus #672 advisory and #660 coexistence boundaries | research_architect, academic-paper intake, pipeline orchestrator |
484| shared/contracts/passport/preregistration_artifact.schema.json | Closed persistent preregistration handoff receipt; companion bytes remain separately named | dispatching layer, intake, pipeline orchestrator |
485| references/systematic_review_toolkit.md | Cochrane v6.4, PRISMA 2020, RoB 2, ROBINS-I, I² guide, GRADE, protocol registration | risk_of_bias, meta_analysis, bibliography, report_compiler |
486| references/literature_monitoring_strategies.md | Google Scholar alerts, PubMed alerts, RSS feeds, Retraction Watch, citation tracking, monitoring cadence | monitoring_agent |
487| references/argumentation_reasoning_framework.md | Cognitive framework for evaluating argument strength: Toulmin model, causal reasoning (Bradford Hill), inference to best explanation, epistemic status classification | synthesis, devils_advocate, source_verification, socratic_mentor, research_architect |
488| references/socratic_mode_protocol.md | Full 5-layer Socratic dialogue flow, management rules, auto-end conditions | socratic_mentor, research_question |
489| references/systematic_review_protocol.md | Full PRISMA pipeline, checkpoint rules, meta-analysis procedures | risk_of_bias, meta_analysis, bibliography, report_compiler |
490| references/cross_agent_quality_definitions.md | Peer-reviewed source tiers, currency standards, severity definitions | all agents |
491| references/changelog.md | Full version history | — |
492
493---
494
495## Templates
496
497| Template | Purpose |
498|----------|---------|
499| templates/research_brief_template.md | Quick mode output format |
500| templates/literature_matrix_template.md | Source x Theme analysis matrix |
501| templates/evidence_assessment_template.md | Per-source quality assessment card |
502| templates/preregistration_template.md | OSF standard 21-item preregistration template |
503| templates/prisma_protocol_template.md | PRISMA-P 2015 systematic review protocol template |
504| templates/prisma_report_template.md | PRISMA 2020 systematic review report template (27 items) |
505
506---
507
508## Examples
509
510| Example | Demonstrates |
511|---------|-------------|
512| examples/exploratory_research.md | Full 6-phase pipeline walkthrough |
513| examples/systematic_review.md | PRISMA-style literature review |
514| examples/policy_analysis.md | Applied comparative policy research |
515| examples/socratic_guided_research.md | Complete Socratic mode multi-turn dialogue (12 rounds) |
516| examples/handoff_to_paper.md | deep-research full mode handoff to academic-paper |
517| examples/review_mode.md | Review mode: 3-agent review pipeline for policy recommendation text |
518| examples/fact_check_mode.md | Fact-check mode: source verification of HEI claims with per-claim verdicts |
519| examples/idea_diversity_coverage_gap_advisory.md | #257 Socratic wording-pattern + lit-review distributional-skew advisories |
520
521---
522
523## Output Language
524
525Follows the user's language. Academic terminology kept in English. Socratic mode uses natural conversational style.
526
527---
528
529## Anti-Patterns
530
531Explicit prohibitions to prevent common failure modes:
532
533| # | Anti-Pattern | Why It Fails | Correct Behavior |
534|---|-------------|-------------|-----------------|
535| 1 | **Confirmation bias in source selection** | Only finding sources that support the hypothesis | Devil's Advocate checkpoint must include counter-evidence search |
536| 2 | **Cherry-picking evidence** | Citing one supportive study while ignoring three contradicting ones | Report the full evidence landscape including conflicting findings |
537| 3 | **Vibe citing** | Mixing elements from 2-3 real papers into a fabricated reference | Every reference must be verified independently; mashup fabrication is the hardest to detect |
538| 4 | **⚠️ IRON RULE: Treating "difficult to verify" as acceptable** | Marking a reference as "uncertain" instead of FAIL | Gray zone = FAIL. If you cannot confirm it exists, it does not go in the report |
539| 5 | **Skipping phases** | Jumping to synthesis before completing source verification | Complete each phase fully; Phase N output is Phase N+1 input |
540| 6 | **Shallow Socratic mode** | Giving answers disguised as questions ("Wouldn't you say X is true?") | Ask genuine questions that expose assumptions; never lead to predetermined conclusions |
541| 7 | **Source tier inflation** | Treating a blog post as equivalent to a peer-reviewed journal | Apply evidence hierarchy strictly: Tier 1 (peer-reviewed) > Tier 2 (preprint) > Tier 3 (gray lit) |
542
543## Quality Standards
544
5451. ⚠️ **IRON RULE**: **Every claim must have a citation** — no unsupported assertions
5462. **Evidence hierarchy** — meta-analyses > RCTs > cohort studies > case reports > expert opinion (field-neutral baseline; grading is **discipline-relative** — a source meeting its own field's gold standard can reach Grade A even at a low design level. See references/source_quality_hierarchy.md §Grading Rubric + §Field-Specific Adjustments)
5473. **Contradiction disclosure** — if sources disagree, report both sides with evidence quality comparison
5484. **Limitation transparency** — every report must have an explicit limitations section
5495. **AI disclosure** — all reports include a statement that AI-assisted research tools were used
5506. **Reproducibility** — search strategies, inclusion criteria, and analytical methods must be documented for replication
5517. **Socratic integrity** — while non-generation Socratic mode is active, never give direct answers; always guide through questions. A candidate response is lawful only after the explicit exit marker and is outside that mode.
552
553## Cross-Agent Quality Alignment
554
555Unified definitions across all agents. ⚠️ IRON RULE: **CRITICAL severity** = issue that would invalidate a core conclusion or constitute academic misconduct. Requires immediate resolution.
556
557> See references/cross_agent_quality_definitions.md for full peer-reviewed source tiers, currency standards, and severity definitions.
558
559---
560
561## Integration with Other Skills
562
563This skill is domain-agnostic but can be combined with domain-specific skills:
564
565```
566deep-research + tw-hei-intelligence -> Evidence-based HEI policy research
567deep-research + report-to-website -> Interactive research report
568deep-research + podcast-script-generator -> Research podcast
569deep-research + academic-paper -> Full research-to-publication pipeline
570deep-research (socratic) + academic-paper (plan) -> Guided research + paper planning
571deep-research (systematic-review) + academic-paper -> PRISMA systematic review paper
572```
573
574---
575
576## Model Tiering (#517, optional)
577
578When ARS_MODEL_TIERING is set, the dispatching session routes this skill's agents per shared/model_tiering.md (canonical: the full 39-agent judgment/execution table + rules). Compact rule:
579
580- **Unset (default):** every agent inherits the session model — byte-equivalent pre-#517 behavior.
581- **economy** (frontier-tier session): execution-type agents dispatch ONE tier below the session model — floor Opus-class, never lower; judgment-type agents stay on the session model. No-op at or below the floor (announce once).
582- **quality-boost** (below-frontier session): judgment-type agents at the checkpoint surfaces (Stage 2.5/4.5 gates; the opt-in Stage 4→5 claim–ref audit; final review) jump UP to the frontier tier (however many tiers away — not a single increment); nothing is ever downgraded. No-op at the frontier (announce once).
583- Unknown values → warn once, behave as unset. Tiers are relative positions, never hard-pinned model ids. When a direction is active, route repeated same-stage calls to the SAME worker so its prompt cache accumulates; unset means dispatch shapes stay byte-equivalent too.
584
585---
586
587## Version Info
588
589| Item | Content |
590|------|---------|
591| Skill Version | 2.12.1 |
592| Last Updated | 2026-08-15 |
593| Maintainer | Cheng-I Wu |
594| Dependent Skills | academic-paper v1.0+ (downstream) |
595
596---
597
598## Version History
599
600> See references/changelog.md for full version history.
601
In the file
SKILL.md4,592 words
Files53
LicenceSource-available
Why you can read it
Nothing in a skill executes. The client loads the text and the model follows it, so a skill can be audited the way a runbook is — by reading it.
What it costs in context
Skills are not billed by the call. They are paid for in context: every token the instructions occupy is a token your code, your diff and your conversation cannot use. Here is what this one takes and when it takes it.
≈300
always loaded
The name and description, so the model knows the skill exists and when to reach for it.
188,225
on trigger
The instruction body and 52 supporting files, read only when the skill fires.
94.3%
of a 200k window
Ten skills this size would take about 943% of the window before you open a file.
050k100k150k200k context window
188.5k tokens, estimated from the bundle at four bytes to the token, held for the rest of the session once it triggers. Heavy. Teams tend to install this one per project rather than globally, and load it only when the job comes up.
Servers bill, skills cost
A server charges by the month. A skill charges once per session, in context, and then keeps charging it for as long as the session lives.
Before and after
The same question, put to the same model twice: once as it comes, and once with these instructions loaded.
No worked example has been published for this skill yet.
Adoption
Installsnone yet
Ratingno reviews yet
The procedure it runs
The procedure has not been published here. It is in the skill’s own SKILL.md, which its author has not sent to the marketplace yet.
Prose, not code
These steps are written for a model to follow, not executed by a runtime. It can still be told to skip one, and it will say so when it does.
Servers it uses
None. This skill calls no MCP servers at all.
Everything it needs is in the instructions, so it works in a project with nothing connected — the model reads the file and changes how it works with what it can already reach.
It writes no files and reaches no network. All it changes is how the model reasons and writes.
What it asks for
Writes filesno
Network accessno
Read from the allowed-tools line of this skill’s own SKILL.md. A skill grants no permissions of its own — it can only ask for tools your client already has.
What it will not do
Every skill is narrow, and the useful ones say where they stop. These are the jobs this one is the wrong tool for.
What this skill is not for has not been published here. Nothing is implied by that: it is a section the author has not filled in.
If this is not it
What is in the bundle
53 files, 754.1 kB on disk. A bundle is text throughout: the instructions the model reads, plus the templates it fills in.
- SKILL.md37.0 kB
- agents/bibliography_agent.md32.8 kB
- agents/devils_advocate_agent.md10.1 kB
- agents/editor_in_chief_agent.md6.8 kB
- agents/ethics_review_agent.md20.8 kB
- agents/meta_analysis_agent.md16.5 kB
- agents/monitoring_agent.md9.4 kB
- agents/report_compiler_agent.md25.3 kB
- agents/research_architect_agent.md25.7 kB
- agents/research_question_agent.md9.9 kB
- agents/risk_of_bias_agent.md11.2 kB
- agents/socratic_mentor_agent.md53.6 kB
- agents/source_verification_agent.md10.4 kB
- agents/synthesis_agent.md27.6 kB
- agents/timeline_extraction_agent.md8.0 kB
- examples/exploratory_research.md7.1 kB
- examples/fact_check_mode.md13.2 kB
- examples/handoff_to_paper.md12.8 kB
- examples/idea_diversity_coverage_gap_advisory.md3.2 kB
- examples/policy_analysis.md9.6 kB
- examples/review_mode.md17.8 kB
- examples/socratic_guided_research.md23.5 kB
- examples/systematic_review.md7.9 kB
- references/apa7_style_guide.md4.7 kB
- references/argumentation_reasoning_framework.md4.0 kB
- references/arxiv_api_protocol.md6.7 kB
- references/changelog.md2.3 kB
- references/chinese_literature_api_protocol.md34.1 kB
- references/cross_agent_quality_definitions.md1.5 kB
- references/crossref_api_protocol.md5.1 kB
- references/equator_reporting_guidelines.md42.8 kB
- references/ethics_checklist.md17.5 kB
- references/failure_paths.md17.6 kB
- references/interdisciplinary_bridges.md21.4 kB
- references/irb_decision_tree.md19.5 kB
- references/literature_monitoring_strategies.md10.8 kB
- references/logical_fallacies.md10.6 kB
- references/methodology_patterns.md17.4 kB
- references/mode_selection_guide.md17.9 kB
- references/openalex_api_protocol.md6.5 kB
- references/preregistration_guide.md13.9 kB
- references/semantic_scholar_api_protocol.md5.2 kB
- references/socratic_mode_protocol.md5.5 kB
- references/socratic_questioning_framework.md11.7 kB
- references/source_quality_hierarchy.md10.4 kB
- references/systematic_review_protocol.md4.0 kB
- references/systematic_review_toolkit.md19.1 kB
- templates/evidence_assessment_template.md3.8 kB
- templates/literature_matrix_template.md3.8 kB
- templates/preregistration_template.md8.7 kB
- templates/prisma_protocol_template.md9.0 kB
- templates/prisma_report_template.md16.2 kB
- templates/research_brief_template.md2.2 kB
What is not in it
No dependencies and nothing executable: a skill is text the agent reads, so the bundle is 53 files you can review in full before installing. The Source-available licence covers the templates and examples as well as the instructions.
Install
Installing copies the bundle into your project. Nothing runs at install time — the files sit on disk until the model reads them.
# Deep Research — Universal Academic Research Agent Team · 188.5k tokens when loaded
npx mcprush@latest skill add imbad0202/deep-research-universal-academic-research-agent-tWrites to .claude/skills/deep-research-universal-academic-research-agent-t/ in the current project. Add --global to put it in your home directory instead, for every project.
Which clients pick it up on their own
A skill is a folder of text. A client with a skills folder reads it without being told; everywhere else the same text works, it is just handed to the model rather than found.
Claude Code.claude/skills/
Claude Desktop
ChatGPT
Cursor.cursor/skills/
VS Code.github/skills/
Codex CLI.agents/skills/
Gemini CLI.gemini/skills/
Grok.grok/skills/
Zed.agents/skills/
Windsurf.windsurf/skills/
Agent SDK.claude/skills/
HTTP / API
This release
Version2.12.1
Publishedno release date on file
PriceFree
Referenceimbad0202/deep-research-universal-academic-research-agent-t
Versions
v2.12.1 is what is on the shelf; no release here carries a date. Instructions change more often than APIs do — a skill can be rewritten entirely without anything it depends on moving.
v2.12.1
- No earlier releases have been published to the marketplace.
Pinning
Put imbad0202/deep-research-universal-academic-research-agent-t@2.12.1 in the install command to hold this exact version. Without the suffix you get whatever is current the day you install, and nothing moves under you afterwards.
Reviews
—no reviews yet · no installs yet
Nobody has reviewed this skill yet. The rating is the mean of the reviews written here, so there is none until somebody writes the first.
Who can post
Only accounts that have had the skill installed for fourteen days, so a review is written after living with it rather than after reading it. Publishers may reply once.
Who wrote it
Publisher
Claim this skillServers0