Academic Paper — Academic Paper Writing Full Pipeline

Workflow

Phase 1: Preparation (Performed Directly by Orchestrator)

1. Extract from user input:
   • Research Topic: Research field, core keywords
   • Research Level (optional): Undergraduate/Master's/Doctoral/Faculty level
   • Target Journal (optional): Journal name, IF range
   • Existing Materials (optional): Data, drafts, literature lists
   • Constraints (optional): Deadline, length, special considerations
2. Create the _workspace/ directory at the project root
3. Organize the input and save it as _workspace/00_input.md
4. If existing files are provided, copy them to _workspace/ and skip the corresponding phase
5. Determine the execution mode based on the scope of the request

Phase 2: Team Assembly and Execution

Inter-Agent Communication Flow:

• designer completes -> delivers variables and sampling plan to experiment, hypotheses and analysis direction to analyst, theoretical background to writer, research field to preparer
• experiment completes -> delivers data structure to analyst, Methods materials to writer
• analyst completes -> delivers Results materials (APA format + Tables/Figures) to writer, analysis code to preparer
• writer completes -> delivers completed manuscript to preparer
• orchestrator cross-validates overall consistency (hypotheses <-> analysis <-> results <-> discussion)

Phase 3: Integration and Final Deliverables

1. Verify consistency across all files in _workspace/
2. Validate logical consistency of hypotheses-analysis-results-discussion chain
3. Generate the integrated review report
4. Report the final summary to the user

Execution Modes by Scope

Using Existing Files: When the user provides data, drafts, or design documents, copy them to _workspace/ and skip the corresponding phase.

Data Transfer Protocol

File naming convention: {order}_{deliverable}.{extension}

Error Handling

Test Scenarios

Normal Flow

Prompt: "Write a paper studying the effect of social media usage time on college students' academic achievement. Survey-based." Expected Results:

• Design: Cross-sectional survey design, social media time (IV) -> GPA (DV), control variables included
• Protocol: Questionnaire (using existing validated scales), online collection procedure
• Analysis: Multiple regression code (R/Python), APA-format results
• Manuscript: Complete IMRaD structure, prior study citations
• Submission: 3 SSCI journal recommendations, cover letter

Existing File Flow

Prompt: "Analyze this data and write the paper" + CSV data attached Expected Results:

• Identify variable structure from the data to develop analysis strategy
• Lightly deploy designer to reverse-extract research framework
• Proceed in order: analyst -> writer -> preparer

Error Flow

Prompt: "Write a paper, but I haven't decided on a topic yet" Expected Results:

• designer asks about the user's area of interest, or
• suggests 3 research topics based on recent trends
• Review report notes "Topic undetermined — proceeded on example basis"

Agent Extension Skills