Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Research Coordinator v12.0 - Human-Centered Edition · brycewang-stanford | DecimalAI

researchby brycewang-stanford

Research Coordinator v12.0 - Human-Centered Edition

brycewang-stanford/research-coordinator

Source brycewang-stanford/Auto-Empirical-Research-Skills ↗License NOASSERTION

Research Coordinator v12.0 - Human-Centered Edition (Systematic Review Automation) Context-persistent platform with 24 specialized agents across 9 categories (A-G, I, X). Features: Human Checkpoints First, VS Methodology, Paradigm Detection, Systematic Review Automation. Supports quantitative, qualitative, mixed methods research, and systematic review automation. Language: English. Responds in Korean when user input is Korean. Triggers: research question, theoretical framework, hypothesis, liter

SkillScore41Maintained

Automatic

Research pack

Last verified 8/8/2026

What it changes

Passed 5 of 12 test cases with the skill, and 2 of 12 without it.

Without skill18%

With skill41%

Checks it passed that the model alone missed

Selects Agent G6 Academic Style Humanizer
Scores divergent options using Typicality T-Scores
Routes the validation task to Agent F5 Humanization Verifier

and 4 more in the full report

6 fixed3 passed either way12 still failing1 broke

Pass-rate lift: +23 pp
Tokens: +176%
Agent turns: 0%
Response time: -26%

Model gemini-3.6-flashMeasured 8/8/2026v19 of 22 cases passed with skill · 12 comparable

Tokens, turns and time describe resource use, not a dollar price.

Use this skill

Save it into your project's skills folder. Claude Code picks it up in its next session.

curl -fsSL --create-dirs -o .claude/skills/brycewang-stanford-research-coordinator/SKILL.md https://app.decimal.ai/s/brycewang-stanford-research-coordinator/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add brycewang-stanford-research-coordinator, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"

1 · One-time — install the CLI (docs ↗)

pip install "decimalai>=0.15"

2 · Get the skill — writes .agents/skills/brycewang-stanford-research-coordinator/ and pins it in .decimal/skills.lock

decimalai skills add brycewang-stanford-research-coordinator

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/brycewang-stanford-research-coordinator/SKILL.md https://app.decimal.ai/s/brycewang-stanford-research-coordinator/SKILL.md

Raw file: /s/brycewang-stanford-research-coordinator/SKILL.md ↗ · machine-readable summary at /s/brycewang-stanford-research-coordinator.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 6 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 6

Prompt

Agent I2 is prepared to run automated AI filtering on 1,200 retrieved paper abstracts. What step is mandatory before execution?

What was checked, without and with the skill
Checked	Without skill	With skill
Enforces the SCH_SCREENING_CRITERIA checkpoint to approve inclusion and exclusion criteria	Missed	Passed

case-14 · gemini-3.6-flash · 8/8/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md294 linessingle file

MANDATORY: Checkpoint Enforcement Rules (v8.2 — MCP-First)

Full details: docs/CHECKPOINT-RULES.md

Rule 5: Override Refusal

사용자가 REQUIRED 체크포인트 스킵 요청 시: → AskUserQuestion으로 Override Refusal Template 제시 (텍스트 거부 아님) → REQUIRED는 어떤 상황에서도 스킵 불가 → 참조: .claude/references/checkpoint-templates.md → Override Refusal Template

Rule 6: MCP-First Verification

에이전트 실행 전: diverga_check_prerequisites(agent_id) 호출 → approved: true → 에이전트 실행 진행 → approved: false → missing 배열의 각 체크포인트에 대해 AskUserQuestion 호출 → MCP 미가용 시: .research/decision-log.yaml 직접 읽기 → 대화 이력은 최후 수단

단일 에이전트 호출 시:

diverga_check_prerequisites(agent_id) 호출
approved: false → 각 missing checkpoint에 대해 AskUserQuestion 도구 호출
REQUIRED 전제조건은 절대 스킵 불가 (사용자가 "건너뛰자"해도 Override Refusal Template 제시)
모든 전제조건 통과 후 에이전트 작업 시작
에이전트 완료 시 diverga_mark_checkpoint() 으로 결정 기록

다중 에이전트 동시 호출 시:

모든 트리거된 에이전트의 prerequisites를 합집합으로 수집
Checkpoint Dependency Order에 따라 정렬 (Level 0 → Level 5)
각 전제조건을 AskUserQuestion 도구로 순서대로 질문
중복 체크포인트는 한 번만 질문
모든 전제조건 해결 후 에이전트들을 병렬 실행
각 에이전트 실행 중 자체 체크포인트도 AskUserQuestion 필수

모든 체크포인트에서:

반드시 AskUserQuestion 도구 사용 (텍스트 질문 금지)
.claude/references/checkpoint-templates.md의 파라미터 사용
응답 받을 때까지 STOP and WAIT
diverga_mark_checkpoint(checkpoint_id, decision, rationale) 으로 결정 기록

자기 검증 (에이전트 작업 완료 전):

"Own Checkpoints"를 모두 트리거했는지 자가 확인

Track record

Benchmark

5 / 12

test cases passed, vs 2 / 12 without

Adoption

Users: 0
Times used: 0

Safety and source details

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/2/2026

View 15 scanner checks

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

brycewang-stanford / Auto-Empirical-Research-Skills

skills/25-HosungYou-Diverga/skills/research-coordinator/SKILL.md

View on GitHub ↗

License NOASSERTIONImported 8/2/2026Version v1

More skills from the same source and nearby in the registry, with their headline benchmark lift.