Linh
Community@xlinh2301
Linh's registry delivers bioinformatics database connectors, autonomous ML research loops, and academic writing pipelines for computational biology and scientific publishing.
Agent Skills by Linh
Showing 85 vetted skills indexed across 1 GitHub repositories.
xlinh2301arbor-agent-tools
Emulates Arbor tree, eval, merge, and report tools with a stdlib Python state script.
xlinh2301human-protein-atlas-database
Retrieve protein expression and subcellular localization data from the Human Protein Atlas API.
xlinh2301research-question
Drafts, scores, and revises research questions against a five-axis rubric with novelty checks.
xlinh2301literature-search-europepmc
Search Europe PMC and download open-access full texts, PDFs, citations, and references.
xlinh2301claim-verify
Verifies each claim in a results draft against its source dataset through reproduction and stress tests.
xlinh2301dbsnp-database
Query NCBI dbSNP to resolve rsIDs, VCF coordinates, and HGVS strings for genetic variants.
xlinh2301purple-team
Orchestrates red-team attack and blue-team fix loops to harden guardrails and classifiers.
xlinh2301protein-sequence-msa
Aligns multiple protein sequences using the EBI Clustal Omega API and reports conservation metrics.
xlinh2301research-proposal
Iteratively grades and revises research proposals against literature-grounded evidence until a passing score.
xlinh2301arbor-agent-orchestrator
Coordinates Arbor-style research runs by loading phase skills for ideation, execution, and merging.
xlinh2301embl-ebi-ols
Query the EMBL-EBI Ontology Lookup Service API for biomedical ontology terms and hierarchies.
xlinh2301interpro-database
Query the InterPro API to identify protein domains, families, and functional annotations.
xlinh2301predictingthepast
Restore, attribute, date, and contextualize ancient Latin and Greek inscriptions using Aeneas and Ithaca models.
xlinh2301jaspar-database
Query the JASPAR API for transcription factor binding profiles and motif matrices.
xlinh2301red-team
Generates adversarial inputs to catalogue distinct failure classes of guardrails and classifiers.
xlinh2301hypothesis-gen
Generates and literature-vets novel research hypotheses through a multi-agent generate-ground-judge loop.
xlinh2301anomaly-investigation
Diagnoses the root cause of a known data anomaly by testing and eliminating candidate explanations.
xlinh2301alphagenome-single-variant-analysis
Analyzes genetic variant effects on gene expression, splicing, and chromatin using the AlphaGenome API.
xlinh2301scientific-writer
Iteratively critiques, grades, and revises scientific drafts using multi-judge review loops.
xlinh2301tournament-autoresearch
Runs an autonomous ML research loop where competing agent-proposed changes are judged before each experiment.
xlinh2301deep-research
Orchestrates a 13-agent pipeline for academic research, literature review, and APA report generation.
xlinh2301plan-loop
Converts engineering prompts into validated, PR-sized task plans with critique-driven revision.
xlinh2301blue-team
Patches failing guardrail or test cases one root-cause class at a time under a regression gate.
xlinh2301literature-search-openalex
Query the OpenAlex scholarly database for papers, authors, institutions, and citation metrics.
Frequently Asked Questions About Linh
FAQPage SchemaWhat tasks can I accomplish with Linh's skill registry?▼
You can query 25+ biomedical databases (UniProt, Ensembl, ChEMBL, ClinVar, PubMed, ClinicalTrials.gov), run autonomous ML research loops that propose and test code changes, verify data-backed claims against datasets, clean tabular data to contracts, and produce publication-ready papers with multi-agent review panels.
Who are these skills designed for?▼
Computational biologists, genomics researchers, ML engineers, and academic authors. Bioinformatics skills target variant analysts and structural biologists; autoresearch loops target ML practitioners optimizing training metrics; the academic pipeline targets researchers preparing manuscripts, rebuttals, and systematic reviews.
What are the runtime requirements and prerequisites?▼
Most analytical loops require Python 3.9+. The swe-loop needs a runnable test command in the target repo; blue-team and purple-team use git plus the gh CLI for pull requests; OpenSpec skills require the openspec CLI; a uv skill installs the uv package manager when other skills need it.
Are these skills open source and what do they cost?▼
The OpenSpec skills carry an explicit MIT license; other skills list no license or fee in their frontmatter. Database skills query free public resources (NCBI, EMBL-EBI, arXiv, OpenAlex), though some loops optionally call paid services like Perplexity Sonar.
How do the red-team, blue-team, and purple-team skills work together?▼
Red-team adversarially probes a guardrail, classifier, or prompt you own and catalogs distinct failure classes. Blue-team patches each failure class under a regression gate and opens a pull request. Purple-team orchestrates both as a closed find-fix-reverify loop until a fresh attack pass surfaces nothing.