Intake-Project Skill
Purpose
This skill is the front door for a new or messy project. It converts a folder, document bundle, or mixed set of notes into a structured project state that other skills can use safely.
Use this skill when:
- a new paper or proposal folder has been created
- an older folder exists but is poorly organized
- the user asks "what is this project and what should I do next?"
- another skill needs a reliable project summary before proceeding
Communication Rules
- Communicate with the user in their preferred language.
- Keep project labels and file names in the language already used by the workspace.
- Use English for manuscript section names, study design names, and medical/statistical terminology.
Inputs
Accept any of the following:
- a project folder
- a manuscript draft
- an abstract or proposal
- tables/figures plus notes
- a mixed folder with PDFs, drafts, and analyses
If information is incomplete, infer cautiously from file names and contents, then label uncertain items clearly.
Core Tasks
1. Project classification
Determine:
- project type:
original | review | meta-analysis | case report | technical note | grant | peer review | challenge | career-doc
- primary domain:
radiology | medical AI | multimodal LLM | intervention | survival/prognostic | diagnostic accuracy | workflow
- target output:
paper | abstract | grant | review | rebuttal | CV
- likely target journal or venue, if recoverable
2. State reconstruction
Identify:
- what already exists
- what is missing
- current phase
- blocking dependencies
3. Project memory scaffold
If missing, propose or create lightweight anchor files:
PROJECT.md
STATUS.md
CLAIMS.md
DATA_DICTIONARY.md
ANALYSIS_PLAN.md
REVIEW_LOG.md
Create only files that are justified by the project type.
4. Action plan
Produce the next 3-5 actions in dependency order.
Canonical Manuscript Folder Structure
For any manuscript project (cohort, MA, RCT, case series), enforce this structure when scaffolding or reorganizing. Map every new artifact into one of these slots — do not invent ad-hoc folders.
{project_root}/
├── HANDOFF.md # session handoff entry point
├── README.md # project overview
├── data/ # raw data (NEVER edit; read-only)
├── analysis/ # reproducible scripts (00_* → 04_*)
├── output/ # analysis outputs: CSVs, PNGs, intermediates
├── irb/ # IRB/ethics docs
├── proposal/ # original protocol / approved proposal
├── reviews/ # external correspondence
├── manuscript/ # SOURCE manuscript + drafting
│ ├── manuscript_v{N}.{md,docx,pdf} # current canonical working version (top level)
│ ├── build_unified_docx.py # or pandoc wrapper
│ ├── archive/ # ALL prior versions v1 .. v{N-1}
│ ├── reviews/ # QC: self_review, peer_review, STROBE/PRISMA, critic
│ ├── figures/ # figure scripts + rendered PNG/PDF
│ └── tables/ # table scripts + rendered docx
└── submission/ # per-journal packages
└── {journal-slug}/ # e.g., chest/, kjr/
├── CHECKLIST.md
├── cover_letter.{md,docx,pdf}
├── title_page.docx # separated for double-anonymized
├── manuscript_anonymized.{docx,pdf}
├── supplement.{docx,pdf}
├── strobe_checklist.md # or PRISMA / CONSORT
├── circulation_email.md
└── figures/ # submission-ready DPI copies
Rules
manuscript/ = source; submission/{journal}/ = derived artifacts. Regenerate submission files from manuscript/manuscript_v{N}.md; never edit anonymized/title-page directly.
- One canonical working version at
manuscript/manuscript_v{N}.{md,docx,pdf}. Older versions move to manuscript/archive/ immediately on version bump.
- No loose files at project root. Only
HANDOFF.md, README.md, folder entries.
- QC artifacts (self_review, peer_review, STROBE, critic reports) live in
manuscript/reviews/, not at manuscript top level.
- On rejection/retarget:
cp -r submission/{old} submission/{new}, then rewrite cover letter and reformat.
- Double-anonymized journals (Chest, AJRCCM): title page and anonymized manuscript MUST be separate files under
submission/{journal}/.
When to apply
- At project intake: scaffold empty structure.
- At first submission prep: create
submission/{journal}/ and populate.
- Mid-project cleanup: when
manuscript/ has >3 versioned files or QC docs at top level, reorganize.
- Before session handoff: reorganize if structure is drifting.
Why it pays off: moving version drafts and QC docs off the manuscript/ top level makes retargeting a different journal a single cp -r submission/<journal> submission/<new_journal>.
Workflow
Phase 1: Discover context
- Read top-level folder names and key files.
- Detect manuscript-like files, tables, figures, protocols, and analysis outputs.
- Extract:
- project title or working title
- study question
- dataset or cohort hints
- collaborators or institutions
- venue/journal hints
Phase 2: Classify project stage
Assign one current stage:
idea
data assembly
analysis planning
analysis in progress
drafting
revision
submission prep
archived/unclear
Gate: Present the classification (project type, stage, target output) to the user.
Confirm before creating any files — misclassification leads to wrong scaffold and
wrong skill routing.
Phase 3: Surface missing inputs
Check for common gaps:
- no explicit study question
- no target journal
- no analysis plan
- no variable dictionary
- no claims-to-results map
- no review log for revised manuscripts
Phase 4: Produce normalized summary
Output this structure:
## Project Intake Summary
Project: ...
Type: ...
Current stage: ...
Likely target: ...
### What exists
- ...
### What is missing
- ...
### Risks / ambiguities
- ...
### Recommended next actions
1. ...
2. ...
3. ...
Optional File Templates
PROJECT.md
# PROJECT
- Title:
- Type:
- Primary question:
- Target journal/venue:
- Lead folder:
- Collaborators:
- Last updated:
STATUS.md
# STATUS
- Current stage:
- Current blocker:
- Next actions:
1.
2.
3.
- Last updated:
Guardrails
- Do not invent data values, outcomes, or collaborator roles.
- Do not assume a target journal unless evidence exists in the files.
- Do not create a large folder scaffold when the user only wants a quick assessment.
- If a project appears to mix multiple studies, say so explicitly rather than collapsing them into one.
Handoff Rules
After intake:
- route to
search-lit if the literature basis is weak
- route to
design-study if the research question exists but design logic is unclear
- route to
manage-project if the folder should be scaffolded
- route to
write-paper only after the project phase is clearly drafting
What This Skill Does NOT Do
- It does not write full manuscript sections
- It does not perform statistical analysis
- It does not verify citations deeply
- It does not replace study design review
Anti-Hallucination
- Never fabricate file paths, URLs, DOIs, or package names. Verify existence before recommending.
- Never invent journal metadata, impact factors, or submission policies without verification at the journal's website.
- If a tool, package, or resource does not exist or you are unsure, say so explicitly rather than guessing.
1---2name: intake-project3description: >4 Intake and normalize a new radiology research project. Classifies project type, summarizes current state,5 identifies missing inputs, recommends next steps, and scaffolds lightweight project memory files.6triggers: new project, intake project, project intake, classify project, organize project, what is this project7tools: Read, Write, Edit, Bash, Grep, Glob8model: inherit9metadata:10 source: https://skillmd.com/skills/aperivue/intake-project11---1213# Intake-Project Skill1415## Purpose1617This skill is the front door for a new or messy project. It converts a folder, document bundle, or mixed set of notes into a structured project state that other skills can use safely.1819Use this skill when:20- a new paper or proposal folder has been created21- an older folder exists but is poorly organized22- the user asks "what is this project and what should I do next?"23- another skill needs a reliable project summary before proceeding2425---2627## Communication Rules2829- Communicate with the user in their preferred language.30- Keep project labels and file names in the language already used by the workspace.31- Use English for manuscript section names, study design names, and medical/statistical terminology.3233---3435## Inputs3637Accept any of the following:38- a project folder39- a manuscript draft40- an abstract or proposal41- tables/figures plus notes42- a mixed folder with PDFs, drafts, and analyses4344If information is incomplete, infer cautiously from file names and contents, then label uncertain items clearly.4546---4748## Core Tasks4950### 1. Project classification5152Determine:53- project type: `original | review | meta-analysis | case report | technical note | grant | peer review | challenge | career-doc`54- primary domain: `radiology | medical AI | multimodal LLM | intervention | survival/prognostic | diagnostic accuracy | workflow`55- target output: `paper | abstract | grant | review | rebuttal | CV`56- likely target journal or venue, if recoverable5758### 2. State reconstruction5960Identify:61- what already exists62- what is missing63- current phase64- blocking dependencies6566### 3. Project memory scaffold6768If missing, propose or create lightweight anchor files:69- `PROJECT.md`70- `STATUS.md`71- `CLAIMS.md`72- `DATA_DICTIONARY.md`73- `ANALYSIS_PLAN.md`74- `REVIEW_LOG.md`7576Create only files that are justified by the project type.7778### 4. Action plan7980Produce the next 3-5 actions in dependency order.8182---8384## Canonical Manuscript Folder Structure8586For any manuscript project (cohort, MA, RCT, case series), enforce this structure when scaffolding or reorganizing. Map every new artifact into one of these slots — do not invent ad-hoc folders.8788```89{project_root}/90├── HANDOFF.md # session handoff entry point91├── README.md # project overview92├── data/ # raw data (NEVER edit; read-only)93├── analysis/ # reproducible scripts (00_* → 04_*)94├── output/ # analysis outputs: CSVs, PNGs, intermediates95├── irb/ # IRB/ethics docs96├── proposal/ # original protocol / approved proposal97├── reviews/ # external correspondence98├── manuscript/ # SOURCE manuscript + drafting99│ ├── manuscript_v{N}.{md,docx,pdf} # current canonical working version (top level)100│ ├── build_unified_docx.py # or pandoc wrapper101│ ├── archive/ # ALL prior versions v1 .. v{N-1}102│ ├── reviews/ # QC: self_review, peer_review, STROBE/PRISMA, critic103│ ├── figures/ # figure scripts + rendered PNG/PDF104│ └── tables/ # table scripts + rendered docx105└── submission/ # per-journal packages106 └── {journal-slug}/ # e.g., chest/, kjr/107 ├── CHECKLIST.md108 ├── cover_letter.{md,docx,pdf}109 ├── title_page.docx # separated for double-anonymized110 ├── manuscript_anonymized.{docx,pdf}111 ├── supplement.{docx,pdf}112 ├── strobe_checklist.md # or PRISMA / CONSORT113 ├── circulation_email.md114 └── figures/ # submission-ready DPI copies115```116117### Rules118119- **`manuscript/` = source; `submission/{journal}/` = derived artifacts.** Regenerate submission files from `manuscript/manuscript_v{N}.md`; never edit anonymized/title-page directly.120- **One canonical working version** at `manuscript/manuscript_v{N}.{md,docx,pdf}`. Older versions move to `manuscript/archive/` immediately on version bump.121- **No loose files at project root.** Only `HANDOFF.md`, `README.md`, folder entries.122- **QC artifacts** (self_review, peer_review, STROBE, critic reports) live in `manuscript/reviews/`, not at manuscript top level.123- **On rejection/retarget:** `cp -r submission/{old} submission/{new}`, then rewrite cover letter and reformat.124- **Double-anonymized journals** (Chest, AJRCCM): title page and anonymized manuscript MUST be separate files under `submission/{journal}/`.125126### When to apply127128- At project intake: scaffold empty structure.129- At first submission prep: create `submission/{journal}/` and populate.130- Mid-project cleanup: when `manuscript/` has >3 versioned files or QC docs at top level, reorganize.131- Before session handoff: reorganize if structure is drifting.132133**Why it pays off:** moving version drafts and QC docs off the `manuscript/` top level makes retargeting a different journal a single `cp -r submission/<journal> submission/<new_journal>`.134135---136137## Workflow138139### Phase 1: Discover context1401411. Read top-level folder names and key files.1422. Detect manuscript-like files, tables, figures, protocols, and analysis outputs.1433. Extract:144 - project title or working title145 - study question146 - dataset or cohort hints147 - collaborators or institutions148 - venue/journal hints149150### Phase 2: Classify project stage151152Assign one current stage:153- `idea`154- `data assembly`155- `analysis planning`156- `analysis in progress`157- `drafting`158- `revision`159- `submission prep`160- `archived/unclear`161162**Gate:** Present the classification (project type, stage, target output) to the user.163Confirm before creating any files — misclassification leads to wrong scaffold and164wrong skill routing.165166### Phase 3: Surface missing inputs167168Check for common gaps:169- no explicit study question170- no target journal171- no analysis plan172- no variable dictionary173- no claims-to-results map174- no review log for revised manuscripts175176### Phase 4: Produce normalized summary177178Output this structure:179180```text181## Project Intake Summary182Project: ...183Type: ...184Current stage: ...185Likely target: ...186187### What exists188- ...189190### What is missing191- ...192193### Risks / ambiguities194- ...195196### Recommended next actions1971. ...1982. ...1993. ...200```201202---203204## Optional File Templates205206### `PROJECT.md`207208```md209# PROJECT210211- Title:212- Type:213- Primary question:214- Target journal/venue:215- Lead folder:216- Collaborators:217- Last updated:218```219220### `STATUS.md`221222```md223# STATUS224225- Current stage:226- Current blocker:227- Next actions:228 1.229 2.230 3.231- Last updated:232```233234---235236## Guardrails237238- Do not invent data values, outcomes, or collaborator roles.239- Do not assume a target journal unless evidence exists in the files.240- Do not create a large folder scaffold when the user only wants a quick assessment.241- If a project appears to mix multiple studies, say so explicitly rather than collapsing them into one.242243---244245## Handoff Rules246247After intake:248- route to `search-lit` if the literature basis is weak249- route to `design-study` if the research question exists but design logic is unclear250- route to `manage-project` if the folder should be scaffolded251- route to `write-paper` only after the project phase is clearly `drafting`252253---254255## What This Skill Does NOT Do256257- It does not write full manuscript sections258- It does not perform statistical analysis259- It does not verify citations deeply260- It does not replace study design review261262## Anti-Hallucination263264- **Never fabricate file paths, URLs, DOIs, or package names.** Verify existence before recommending.265- **Never invent journal metadata, impact factors, or submission policies** without verification at the journal's website.266- If a tool, package, or resource does not exist or you are unsure, say so explicitly rather than guessing.