agentleFS
jxtsejxtse/scientific-research-skills/skills/related-work-survey/SKILL.md

related-work-survey

Skill68 starsChanged 7 months agoMITView on GitHub

Conducts a systematic related-work / literature-survey / state-of-the-art review for a research question by defining survey dimensions, searching each axis, building a taxonomy of prior work, identifying the gap, and producing a positioning narrative for a paper's Related Work section. Goes beyond a flat paper list to deliver structured analysis. Use when the user is starting a new research project and needs to map the landscape, asks "what's been done on X?" or "how does my idea compare to existing work?", needs to write or revise a Related Work / Background / Prior Art section, wants to identify a research gap or position their contribution, or asks to build a taxonomy of approaches in a research area.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every research topic sits at an intersection of multiple dimensions.

    line 38
  2. Distinctive:

    Be honest about overlap: If someone has done something very similar, acknowledge it

    line 124

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Related Work Survey

Systematic literature survey for positioning a research contribution.

When to Use

  • User starts a new research project and needs landscape understanding
  • User asks "what's been done on X?"
  • User needs to write a related work section
  • User wants to identify the gap their work fills

Workflow

Step 1: Define the Research Question

Work with the user to pin down:

  • Core question: What specific problem are we solving?
  • Key concepts: What are the 3-5 key terms/concepts?
  • Scope boundaries: What's in scope vs. adjacent but out of scope?

Step 2: Identify Survey Dimensions

Every research topic sits at an intersection of multiple dimensions. Identify 2-4 axes:

Example for "Neural-Symbolic Decomposition in LLM Agents":

  • Axis 1: Neuro-symbolic integration approaches
  • Axis 2: LLM agent architectures (harness, scaffolding)
  • Axis 3: Adaptive/metacognitive planning
  • Axis 4: Text analysis at scale (specific application)

Step 3: Search Each Dimension

For each axis, use literature-search skill with targeted queries:

Axis 1 → "neuro-symbolic integration LLM reasoning 2024 2025"
Axis 2 → "LLM agent harness scaffolding architecture survey"
Axis 3 → "metacognitive planning adaptive tool use LLM"
Axis 4 → "large scale text analysis LLM code generation quality"

Collect 10-20 papers per axis, then deduplicate across axes.

Step 4: Build the Taxonomy

Organize papers into a structured taxonomy:

## Related Work Taxonomy

### 1. Neuro-Symbolic Integration
  1.1 Neural reasoning with symbolic verification
  1.2 Symbolic planning with neural execution
  1.3 Adaptive decomposition (our focus)

### 2. LLM Agent Architecture
  2.1 Harness engineering
  2.2 Self-evolving agents
  2.3 Meta-Harness optimization

### 3. ...

Step 5: Identify the Gap

The gap is where your work lives — the intersection that no existing paper covers:

Paper A does X but not Y.
Paper B does Y but not X.
We do both X and Y, connected by Z.

Step 6: Write the Narrative

Structure the related work section as a story:

  1. Context: What's the broader field?
  2. Prior art per axis: What has been done?
  3. Gap statement: What's missing?
  4. Our position: How do we fit in?

Output Format

Deliver a structured document:

# Related Work Survey: [Topic]

## Research Question
[1-2 sentences]

## Taxonomy
[Structured tree]

## Key Papers
| Paper | Axis | Key Contribution | Gap Relative to Us |
|-------|------|------------------|--------------------|
| ...   | ...  | ...              | ...                |

## Identified Gap
[Clear statement of what no existing work covers]

## Recommended Positioning
[How to frame our contribution relative to existing work]

Principles

  • Be honest about overlap: If someone has done something very similar, acknowledge it
  • Recent > old: Prioritize 2024-2026 papers for positioning; cite older seminal works for foundations
  • Quality > quantity: 20 well-chosen papers > 50 tangentially related ones
  • Track the conversation: Note who cites whom — citation chains reveal intellectual lineage
  • Check workshops/findings: Important early-stage work often appears in workshop papers

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.