/research-mission-generator

name: research-mission-generator
description: "Package conversation research into a GSD-ready mission package. Produces a three-stage Vision → Research → Mission pipeline as a LaTeX PDF following GSD/NASA SE methodology, designed to be handed to gsd-skill-creator for execution. Use this skill when the user has been discussing, researching, or brainstorming a topic and then asks to turn it into a research mission, research pack, mission package, or says 'package this as a mission', 'make this a research pack', 'turn this into a mission for skill-creator', 'create a research mission from this', or 'use the research mission skill'. The skill harvests findings, sources, and structure from the current conversation and any prior research, then produces the complete pipeline document. Also trigger if the user asks to 'create a research mission on [topic]' cold — in that case, conduct web research first, then package."
version: 1.0.0
format: 2025-10-02
triggers:
  - "the user has been discussing, researching, or brainstorming a topic and then asks to turn it into a research mission, research pack, or mission package, or says 'package this as a mission'"
  - "if the user asks to 'create a research mission on [topic]' cold — in that case, conduct web research first, then package"
updated: 2026-04-25
status: ACTIVE

Research Mission Generator

Package conversation research into publication-quality GSD mission packages. Produces a three-stage pipeline (Vision → Research → Mission) as a LaTeX PDF, ready for handoff to gsd-skill-creator.

Primary Workflow

The typical pattern is:

1. **User has a conversation** — discussion, brainstorming, web research, thinking time, or any exploration of a topic 2. **User says "turn this into a research mission"** — triggering this skill 3. **Skill harvests the conversation** — extracting findings, sources, structure, and insights already developed 4. **Skill fills gaps** — conducting additional web research only where the conversation left holes 5. **Skill produces the package** — three-stage LaTeX PDF + .tex source + index.html 6. **User hands package to gsd-skill-creator** — for execution via Claude Code

The skill is a *structuring and packaging* tool, not a from-scratch generator. The intellectual work has already happened in conversation. This skill gives it bones.

When to Use

**Primary triggers (conversation already happened):**

• "Turn this into a research mission"
• "Package this as a mission for skill-creator"
• "Make this a research pack"
• "Use the research mission skill on what we've discussed"
• "Create a mission package from this conversation"
• "Structure this for GSD"

**Secondary trigger (cold start — research needed first):**

• "Create a research mission on [topic]"
• "I want a deep research study into [topic]"
• In this case, conduct 3–6 web searches first (see Step 2b below), then proceed with packaging.

Workflow

Step 1: Harvest the Conversation

Scan the current conversation (and any referenced past conversations) to extract:

| Element | Where to Find It | Fallback | |---------|-------------------|----------| | **Topic & scope** | User's original question or framing | Ask user | | **Key findings** | Research results, web search outputs, discussion points | Conduct web research | | **Sources cited** | URLs, papers, organizations mentioned in conversation | Validate via web search | | **Structure** | Natural groupings that emerged in discussion | Infer modules from topic | | **Quantitative data** | Numbers, percentages, measurements discussed | Source from research | | **Sensitivity areas** | Cultural, safety, or ethical concerns raised | Detect from domain | | **Insights & connections** | Cross-topic relationships the user identified | Preserve in synthesis module | | **User's framing** | Metaphors, analogies, philosophical angles | Carry into vision narrative and through-line |

**Critical:** Preserve the user's own insights, framings, and connections. The conversation is the primary source material. Don't replace the user's thinking with generic content — amplify it with structure.

Step 2a: Gap Analysis (Conversation-Based)

After harvesting, identify what's missing for a complete three-stage package:

• **Vision gaps:** Is there a clear problem statement? Are there enough concrete problems (need 3–5)?
• **Research gaps:** Are all claims sourced? Are there modules with thin evidence?
• **Mission gaps:** Can the topic decompose into 4–6 parallelizable modules?

Conduct targeted web searches ONLY to fill identified gaps. Don't re-research what the conversation already covered well.

Step 2b: Cold-Start Research (Topic-Only)

If the user provides just a topic without prior conversation, conduct full research:

1. **Broad overview** — `[topic] overview` 2. **Authoritative sources** — `[topic]` + relevant agency names 3. **Peer-reviewed findings** — `[topic] research study findings` 4. **Subtopic deep-dives** — One search per identified module 5. **Threats/challenges** — `[topic] threats challenges risks` 6. **Unique/endemic features** — `[topic] unique rare endemic`

**Source quality rules (ABSOLUTE):**

• Government agencies (USGS, USFS, NPS, NOAA, NIH, NASA, EPA, etc.)
• Peer-reviewed journals
• University research programs
• Professional organizations
• NEVER entertainment media, blogs, or unsourced claims

Step 3: Identify Research Modules

Decompose into 4–6 research modules. Prefer modules that emerged naturally from conversation structure. Each module must be:

• **Self-contained** — An agent can produce it from its spec alone
• **Parallelizable** — At least 2 modules can run simultaneously
• **Connected** — At least one cross-module relationship exists

Common module patterns by domain:

| Domain | Typical Modules | |--------|----------------| | Ecology/Biology | Flora, Fauna, Fungi/Microbiome, Aquatic, Ecological Networks | | Technology | Architecture, Implementation, Security, Performance, Integrat