Task Recipes

Use this page when you know your real-world goal, but you do not yet know which stages, Task IDs, or skills to lean on.

This page is intentionally task-first:

• start from the job you need done
• map it to stages and Task IDs
• understand which skills are typically involved
• choose the smallest route that still gives you defensible outputs

If you want the full stage-by-stage map, use Skills Guide. If you want paper-type defaults such as "systematic review" or "methods paper," use Examples.

How To Use This Page

For each scenario below, read it in this order:

1. When to use
2. Minimal route
3. Deeper route
4. Typical skills
5. Typical outputs

Do not assume you must run every stage. The best workflow is usually the narrowest route that still satisfies your paper type and evidence needs.

1. I Need To Turn A Broad Topic Into A Researchable Question

When to use

Use this when the topic is still fuzzy, the contribution is not yet clear, or the venue target is still moving.

Minimal route

• A1: refine the question and define scope
• A4: identify the strongest gap

Deeper route

• A1
• A1_5: generate hypotheses or propositions
• A2 / A3: build theory and positioning
• A5: check venue fit

Typical skills

• question-refiner
• hypothesis-generator
• theory-mapper
• gap-analyzer
• venue-analyzer

Typical outputs

• refined research question set
• contribution framing
• theory map
• prioritized gap memo
• venue-fit constraints

Good first command

python3 -m bridges.orchestrator task-plan \
  --task-id A1 \
  --paper-type empirical \
  --topic ai-in-education \
  --cwd .

2. I Need A Defensible Related-Work Or Systematic Review Base

When to use

Use this when the main bottleneck is corpus quality, screening discipline, extraction consistency, or PRISMA-style transparency.

Minimal route

• B1_5: concept and keyword expansion
• B2: focused paper reading / extraction
• B3: literature mapping

Deeper route

• B1: full search
• B1_5
• B2
• B3
• B4 / B5: citation expansion and/or synthesis support
• G1: PRISMA check before submission

Typical skills

• academic-searcher
• concept-extractor
• paper-screener
• paper-extractor
• citation-snowballer
• literature-mapper
• prisma-checker

Typical outputs

• search log
• screening log
• extraction table
• literature map
• PRISMA-ready counts and compliance memo

Good first command

python3 -m bridges.orchestrator task-run \
  --task-id B2 \
  --paper-type systematic-review \
  --topic ai-in-education \
  --cwd . \
  --research-depth deep

3. I Need To Design An Empirical Study Before Writing Or Coding

When to use

Use this when your question is already stable, but the design, variables, robustness plan, or dataset path is still weak.

Minimal route

• C1: main design
• C2 or C3: variables / data feasibility

Deeper route

• C1
• C1_5 / C2: rival explanations and variable logic
• C3: dataset feasibility
• C3_5 / C4: robustness and data management
• C5: prereg-style handoff

Typical skills

• study-designer
• rival-hypothesis-designer
• dataset-finder
• variable-constructor
• robustness-planner

Typical outputs

• design spec
• analysis plan
• variable specification
• dataset plan
• robustness plan

Good first command

python3 -m bridges.orchestrator task-run \
  --task-id C1 \
  --paper-type empirical \
  --topic ai-in-education \
  --cwd .

4. I Need To Turn Results Into A Manuscript

When to use

Use this when the analysis exists, but the problem is story structure, tables, figures, interpretation, or abstract/title quality.

Minimal route

• F1: outline or manuscript architecture
• F2: paragraph- or section-level writing

Deeper route

• F1
• F2
• F3: full manuscript drafting
• F4: tables/figures/results interpretation support
• F5 / F6: abstract, title, keyword, and final polishing support

Typical skills

• manuscript-architect
• analysis-interpreter
• effect-size-interpreter
• table-generator
• figure-specifier
• meta-optimizer

Typical outputs

• manuscript outline
• section drafts
• result narratives with uncertainty preserved
• paper-ready tables and figure specifications
• optimized title/abstract/keywords

Good first command

python3 -m bridges.orchestrator task-run \
  --task-id F3 \
  --paper-type empirical \
  --topic ai-in-education \
  --cwd . \
  --focus-output manuscript/manuscript.md \
  --output-budget 2

5. I Need Academic Code, Not Generic Product Engineering

When to use

Use this when the work is a methods paper, empirical pipeline, reproducibility package, or statistics-heavy implementation.

Minimal route

• I5: specification
• I6: zero-decision plan

This is the right choice when you are still locking constraints before writing code.

Deeper route

• I5
• I6
• I7: implementation and profiling
• I8: code/statistical review
• I4: reproducibility audit

Typical skills

• code-specification
• code-planning
• code-execution
• code-review
• reproducibility-auditor
• stats-engine

Typical outputs

• code/code_specification.md
• code/plan.md
• code/performance_profile.md
• code/code_review.md
• code/reproducibility_audit.md

Good first command

python3 -m bridges.orchestrator code-build \
  --method "Staggered DID" \
  --topic policy-effects \
  --domain economics \
  --focus full \
  --paper-type methods \
  --cwd .

If outputs are too broad

Use:

• --only-target for selective rerun
• --research-depth deep when reasoning is too shallow
• --focus-output and --output-budget when artifact spread is too wide

6. I Need To Stress-Test, Rebut, Or Package For Submission

When to use

Use this when the manuscript is near completion, already under review, or needs a pre-submission harsh check.

Minimal route

• H1: submission package
• H2: rebuttal support

Deeper route

• G1 / G2: reporting and PRISMA checks
• G4: tone cleanup
• H1
• H2
• H3: peer-review simulation
• H4: fatal-flaw scan

Typical skills

• submission-packager
• rebuttal-assistant
• peer-review-simulation
• fatal-flaw-detector
• reviewer-empathy-checker
• reporting-checker

Typical outputs

• cover letter and submission bundle
• point-by-point rebuttal matrix
• simulated reviewer report
• fatal-flaw memo
• response-tone adjustment log

Good first command

python3 -m bridges.orchestrator task-run \
  --task-id H3 \
  --paper-type empirical \
  --topic ai-in-education \
  --cwd . \
  --triad

7. I Only Need A Narrow Follow-Up, Not A Full Rerun

When to use

Use this when you already have Stage-I artifacts and only need to rerun a few plan steps or fix a small set of review findings.

Typical route

• task-run --only-target <target>
• code-build --only-target I6:S1
• code-build --only-target I8:P1-01

Good examples

python3 -m bridges.orchestrator task-run \
  --task-id I6 \
  --paper-type methods \
  --topic policy-effects \
  --cwd . \
  --only-target S1

python3 -m bridges.orchestrator code-build \
  --method "Staggered DID" \
  --topic policy-effects \
  --domain economics \
  --focus full \
  --only-target I8:P1-01 \
  --cwd .

Which Page Should You Use Next?

• Need the global map of stages and skills: Skills Guide
• Need paper-type defaults: Paper Type Playbooks
• Need exact command flags: CLI Reference
• Need collaboration/runtime details: Agent + Skill Collaboration
• Need installation or upgrade help: Guide Home