- Home
- Skills
- Technical Documentation
- research-gap
Works with the AI Agents you already use
Research Gap
Systematically identifies research gaps and unexplored dimensions across six technical domains. Find the questions nobody answered, analyze your papers and uncover the gaps the community missed.
$7
research-gap
Example session with this skill installed
Analyzed 24 papers (2018–2025) across segmentation methods and found 5 concrete gaps, ranked by impact and difficulty. The most promising: generalization to out-of-distribution scanners and domain shift.
- Read your context and instructions
- Compiled the research-gap
I have analyzed the provided 15 papers on GNNs for fraud detection. The community focuses heavily on static accuracy (AUC/F1) but largely ignores temporal decay and adversarial robustness.
Identified research gaps
| Gap | Why it matters | Difficulty | Impact |
|---|---|---|---|
| Cross-scanner generalization | Models degrade 15–30% on unseen scanners/centers | High | High |
| Few-shot organ segmentation | Rare anatomies lack labeled data | Medium | High |
| Uncertainty-aware segmentation | Point estimates hide model confidence | Medium | Medium |
| Multi-modal fusion (CT+MRI+text) | Single-modality limits clinical use | High | High |
| Explainability for clinicians | Trust blocks deployment | Low | Medium |
Where the community is stuck
- Missing datasets: no large public benchmark with multi-center, multi-scanner labels for abdominal organs.
- Incomplete comparisons: most papers benchmark on single-center data (e.g., BTCV, LiTS) and don't report domain-shift performance.
- Under-explored: test-time adaptation for segmentation is rarely combined with foundation models (e.g., SAM).
Recommended direction
Few-shot, cross-scanner segmentation with a foundation-model backbone — combines the two highest-impact gaps and has no dominant existing solution.
Next steps
- Run
novelty-analysisto validate this direction against prior work. - Run
methodology-designerto plan the benchmark and baselines.
Connects securely to your tools. The creator never sees your data.
About this skill
The problem
The difference between a rejected paper and a published one is usually the gap it fills. This skill finds it systematically.
What it does
- Analyzes your paper collection across six unexplored dimensions: problems, methods, data, evaluation, theory, and domain.
- Surfaces open problems, missing datasets, and under-explored research questions.
- Flags incompletely compared methods, missing benchmarks, and unreproduced results.
- Ranks gaps by impact vs. feasibility and recommends concrete research directions.
- Identifies unexplored problem variants and real-world edge cases existing methods fail to address.
- Analyzes methodological combinations and cross-disciplinary techniques not yet applied to the domain.
- Maps missing datasets, modality gaps, and benchmark limitations across the current literature.
- Detects evaluation flaws, including missing head-to-head comparisons and unmeasured performance metrics like robustness or fairness.
- Prioritizes identified gaps into a feasibility-versus-impact matrix for strategic research planning.
Why this beats prompting it yourself
Generic prompts often return vague suggestions like "more data is needed." This skill uses a six-dimension framework to systematically interrogate the research space, ensuring technical gaps like theoretical guarantees and scale limitations are surfaced with specific evidence.
Use cases
- Defining a PhD thesis topic or a high-impact research direction.
- Writing the "Problem Statement" or "Research Gaps" section of a manuscript.
- Validating the novelty of a proposed contribution before starting implementation.
- Identifying "low-hanging fruit" for quick publication through missing method combinations.
Known limitations
Accuracy depends entirely on the quality and breadth of the provided paper collection. It cannot identify gaps in work that has not been published or indexed.
Important
It identifies gaps; it does not validate a specific idea as novel. Pair it with novelty-analysis for that.
Why buy it
It turns "what's been done" into "what I should do." Use it right after literature-discovery or state-of-the-art to land your contribution in clear, unclaimed territory.
Analysis quality scales with the depth of the provided paper collection. Cannot account for concurrent work in progress or unpublished preprints not included in the input.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 2 months ago
Permissions
No special permissions declared or detected