# AI use cases for Scientific Agent Skills (K-Dense-AI/scientific-agent-skills) Repo: [K-Dense-AI/scientific-agent-skills](https://github.com/K-Dense-AI/scientific-agent-skills) | Use cases: 4 | Page: https://dedupedai.com/use-cases/k-dense-ai-scientific-agent-skills/ A library of 177 Agent Skills that give AI agents documented, validated guidance for scientific packages, databases, and research workflows. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Pull compound and protein records with source provenance: https://dedupedai.com/use-cases/k-dense-ai-scientific-agent-skills/look-up-compound-and-protein-records/ - Screen a compound list for drug-like properties: https://dedupedai.com/use-cases/k-dense-ai-scientific-agent-skills/screen-compounds-for-drug-likeness/ - Run a single-cell RNA-seq analysis with Scanpy: https://dedupedai.com/use-cases/k-dense-ai-scientific-agent-skills/analyze-single-cell-rna-seq-data/ - Draft a literature review with traceable citations: https://dedupedai.com/use-cases/k-dense-ai-scientific-agent-skills/draft-a-sourced-literature-review/ ## How to pull compound and protein records with source provenance with Scientific Agent Skills Repo: [K-Dense-AI/scientific-agent-skills](https://github.com/K-Dense-AI/scientific-agent-skills) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Query public biology and chemistry databases and get each record with its source endpoint. **Who it is for:** A pharmacologist or research analyst who needs compound and protein records for a target list and must show where each value came from. **The problem:** Gathering identifiers and properties from several databases by hand is slow, and records often lack a clear source trail. **The outcome:** A table of records from the named databases, where every row lists the endpoint and access date used. ### Steps 1. `git clone https://github.com/K-Dense-AI/scientific-agent-skills` (inferred, check the README) 2. Follow the install steps in the Getting Started section of the README for your agent host. 3. Open skills/database-lookup/SKILL.md, which documents 80 databases with endpoint selection, pagination, and provenance. 4. Give your agent the target list and the databases to query, such as PubChem, ChEMBL, or UniProt. 5. Check that each row in the returned table lists its source endpoint before you use the data. ### Prompt to paste into your agent ```text Using the scientific-agent-skills repo, read skills/database-lookup/SKILL.md and query PubChem, ChEMBL, and UniProt for . Return a table of identifiers and key properties, with the endpoint and access date for each row. Verify that every row names its source endpoint and flag any database that needs a key you do not have. ``` ### Caveat The lookup skill covers 80 databases with different access rules, so check each database's terms before you use its data. Risk, credentials: Some databases need registered or licensed access, so store API keys outside shared notes and project files. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/k-dense-ai-scientific-agent-skills/look-up-compound-and-protein-records/ ## How to screen a compound list for drug-like properties with Scientific Agent Skills Repo: [K-Dense-AI/scientific-agent-skills](https://github.com/K-Dense-AI/scientific-agent-skills) | Difficulty: Advanced | Time: 1 hour | Basis: From the README Use RDKit-based skills to predict molecular properties and flag ADMET concerns in a candidate set. **Who it is for:** A medicinal chemist with a list of candidate molecules who needs a first-pass property filter before wet-lab work. **The problem:** Checking many structures one at a time for properties and ADMET flags takes time and is easy to do inconsistently. **The outcome:** A ranked table of candidates with computed properties and notes on ADMET concerns, with the code used to compute them. ### Steps 1. `git clone https://github.com/K-Dense-AI/scientific-agent-skills` (inferred, check the README) 2. Follow the install steps in the Getting Started section of the README for your agent host. 3. Find the cheminformatics skill folder under skills/ that covers RDKit and molecular property prediction. 4. `pip install rdkit` (inferred, check the README) 5. Give your agent a file of SMILES strings and ask for property and ADMET screening. ### Prompt to paste into your agent ```text Using the scientific-agent-skills repo and its RDKit cheminformatics skill, compute molecular properties and ADMET flags for the SMILES in . Return a ranked table with the computed values and the code used. Verify that every input structure parses and that no row is silently dropped. ``` ### Caveat The README lists these screening topics but does not describe a full pipeline, so the agent must build and check the workflow itself. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/k-dense-ai-scientific-agent-skills/screen-compounds-for-drug-likeness/ ## How to run a single-cell RNA-seq analysis with Scanpy with Scientific Agent Skills Repo: [K-Dense-AI/scientific-agent-skills](https://github.com/K-Dense-AI/scientific-agent-skills) | Difficulty: Advanced | Time: 1 hour | Basis: From the README Have an agent follow the Scanpy skill to process a single-cell dataset and report clusters. **Who it is for:** A bench biologist or bioinformatics staff member with a single-cell RNA-seq dataset and no time to write the pipeline from scratch. **The problem:** Writing a single-cell pipeline from scratch takes days of package reading and debugging. **The outcome:** A reproducible notebook or script that loads the data, filters cells, clusters them, and reports the main groups found. ### Steps 1. `git clone https://github.com/K-Dense-AI/scientific-agent-skills` (inferred, check the README) 2. Follow the install steps in the Getting Started section of the README for your agent host. 3. Locate the single-cell and Scanpy skill folder under skills/. 4. `pip install scanpy` (inferred, check the README) 5. Give your agent the dataset path and ask for quality control, clustering, and a cluster summary. ### Prompt to paste into your agent ```text Using the scientific-agent-skills repo and its Scanpy single-cell skill, process the dataset at with quality control, clustering, and a cluster summary. Return a script and a short report of cell counts per cluster. Verify the cell and gene counts before and after filtering. ``` ### Caveat The README names Scanpy among the package skills but does not specify input formats, so confirm your file type before the run. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/k-dense-ai-scientific-agent-skills/analyze-single-cell-rna-seq-data/ ## How to draft a literature review with traceable citations with Scientific Agent Skills Repo: [K-Dense-AI/scientific-agent-skills](https://github.com/K-Dense-AI/scientific-agent-skills) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Synthesize papers on a topic and keep each claim linked to a citable source. **Who it is for:** A graduate student or principal investigator preparing a related-work section who needs claims tied to real sources. **The problem:** Literature summaries often drift from the sources, and citations get lost during drafting. **The outcome:** A draft review where each claim points to a source, with a bibliography ready for a reference manager. ### Steps 1. `git clone https://github.com/K-Dense-AI/scientific-agent-skills` (inferred, check the README) 2. Follow the install steps in the Getting Started section of the README for your agent host. 3. Find the scientific communication skills for evidence-traceable writing, literature synthesis, and citation management under skills/. 4. Give your agent a list of papers or a search topic and a target length. 5. Read each claim against its source before you submit the draft. ### Prompt to paste into your agent ```text Using the scientific-agent-skills repo and its evidence-traceable writing and literature synthesis skills, write a literature review on from the papers listed in . Return a draft with a citation after each claim and a matching bibliography. Verify that every cited claim appears in the named source. ``` ### Caveat The README describes these writing skills but does not say how the agent reads paper full text, so supply the papers yourself and check the citations. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/k-dense-ai-scientific-agent-skills/draft-a-sourced-literature-review/