LibreGene MCP server
Commit Nov 15, 2026Lets an AI agent perform plasmid editing, primer design, restriction analysis, and sequence alignment through an integrated MCP server.
Also known as dl-li/LibreGene
- Stars
- 5
- Tools
- —
- Downloads, 30 days
- —
- Latest version
- —
- Last commit
- Nov 15, 2026
- License
- GPL-3.0
What is LibreGene MCP?
LibreGene lets an AI agent perform plasmid editing, primer design, restriction analysis, and sequence alignment through an integrated MCP server.
Summary generated from the project's README
About LibreGene MCP server
Molecular biologists and genetic engineers can manage complex DNA workflows by connecting an AI agent to the LibreGene desktop application. This integration allows an LLM to operate an open plasmid editor as if it were a human user, with changes reflecting live in the user interface. The connection is established via a local loopback server using Bearer-token authentication, which is configured through an auto-generated setup prompt provided within the application.
The server exposes 16 specialized tools that cover the entire genetic engineering lifecycle. Users can perform sequence editing, feature annotation, and codon optimization for nine different species. The agent can also handle primer design for fragment amplification or site-directed mutagenesis, find restriction sites using a database of over 900 enzymes, and manage sequence alignments with Sanger chromatogram data. This setup transforms a standard desktop editor into an interactive environment where an AI can execute multi-step biological design tasks.
What you can do with it
- 1Automated Primer DesignAn agent can design cloning primers with specific restriction tails and select verification primers for colony-PCR.
- 2Sequence Alignment and Mutation AnalysisA developer can use the agent to compare Sanger reads against a reference to identify successful mutations.
- 3Codon Optimization for ExpressionThe agent can convert sequences to match the codon usage of a specific expression host.
- 4RNAi Plasmid ConstructionAn AI can design a plasmid targeting a specific gene based on an experimental protocol and antisense sequence.
Written from the project's README. Every use case is checked against a quote from it.
Automated checks
Ran Nov 15, 2026Full report →Not verified yet. A server needs a score of 80, no serious finding, and enough data for us to check. How checks work · What to fix for more points
- Source repository linked and reachabledl-li/libregene is public and reachablegithub.com/dl-li/libregene ↗Pass
- Package names this repository as its sourceNot checked
- Build provenance (Sigstore attestation)Not checked
- Open source license (OSI)GPL-3.0spdx.org/licenses/GPL-3.0.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-11-13github.com/dl-li/libregene ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNot checked
- No critical advisory in the package or its dependenciesNot checked
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeNot checked
README
View on GitHubA lightweight cross-platform desktop plasmid editor. Feature annotation, primer design, restriction analysis, sequence alignment, RNA folding — all free and open-source.
Features
Feature Annotation
- Clean, minimal display — CDS and mRNA features are translated on the fly, amino acids rendered right under the sequence
- Auto-detection — paste a sequence into the New Sequence dialog and common features (promoters, resistance markers, ori, tags…) are detected automatically and can be annotated with one click
- ORF search — scan open reading frames on both strands, all six frames, displayed inline
- GC content track — sliding-window GC% rendered as a gradient band below the sequence, with adjustable window size
- Compound (multi-segment) features, custom colors, strand toggling, enriched GenBank I/O that preserves colors and primers
Primers
- Binding visualization — mismatches, annealing region and Tm shown at a glance for every primer
- Assisted design — fragment amplification, overlap-extension PCR (OE-PCR) and site-directed mutagenesis, with candidates ranked by Tm / GC% and optional restriction-site tails
Codon Optimization
- 9 species codon usage tables — optimize any CDS/mRNA feature for the expression host of your choice
- Three strategies — use best codon, match codon usage, or harmonize against the source species
- Preview before applying — translation check, CAI and GC% before/after at a glance; optionally avoid creating specified restriction sites
Restriction Enzymes
- Built-in 900+ enzyme database with methylation-aware filtering
- Clear categorization — unique cutters, twice cutters, blunt ends and Type IIS enzymes are distinguished visually, and cut sites are marked on the feature scrollbar for quick navigation
- Custom enzyme sets — define your own enzyme collections and switch between them
- Provider detail data — per-enzyme buffer compatibility, incubation temperature, methylation sensitivity and inactivation info from New England Biolabs, Thermo Fisher FastDigest and Yugong Biotech (BestEnzymes)
Digest and Run Gel
- In-silico digest simulation — pick enzymes, set cutting efficiency, and enumerate every resulting fragment, including partial-digest products
- Virtual gel electrophoresis — agarose/PAGE matrix, concentration and run-time controls with built-in or custom ladders; topology-aware migration (uncut circles run as supercoiled bands) and band diffusion are modeled
Sequence Alignment
- Multiple formats — import Sanger reads (.ab1), FASTA or GenBank sequences and align them against the reference
- Mismatches, insertions and deletions highlighted in place, per-read identity computed automatically
- Sanger chromatogram display — open a
.ab1file to see its trace curves alongside the sequence
RNA Secondary Structure
- MFE folding — minimum-free-energy secondary structure prediction, computed locally
- Interactive layout — force-directed structure view with pan/zoom and draggable nucleotides; can be shown as the editor background
MCP / LLM Agent Integration
LibreGene embeds an MCP server (loopback only at 127.0.0.1:8766, Bearer-token auth) so an LLM agent in your terminal can operate the open plasmid like a real user — while the UI updates live. Launch LibreGene, open the MCP Server dialog in the sidebar (or the "Connect an LLM agent via MCP" link on the empty screen), and copy the auto-generated setup prompt — with the URL and access token embedded — to your agent; it configures the connection itself.
16 tools are exposed, covering the full editing workflow:
- Projects & files —
open_project,save_file(with subsequence/region export),list_workspace,add_to_workspace - Reading —
read_sequence,get_project_overview,get_region_view - Editing —
edit_sequence(insert/delete/replace),set_feature(create & update) - Primers —
add_primer,test_primers(ad-hoc binding checks with Tm and amplicon size),design_primers(amplify / OE-PCR / mutagenesis) - Analysis —
find_restriction_sites,find_orfs,add_alignment,convert_sequence(sequence conversion + codon optimization for 9 species)
Example Tasks
Ready-to-run agent tasks live in examples/tasks, each with a Prompt.txt you can hand to your agent and a reference result:
- Primer Design — clone mEGFP into a BamHI/HindIII-digested BlueScribe vector: design cloning primers with restriction tails, pick colony-PCR verification primers, then design A206K mutagenesis primers
- Alignment — given pVA-MCS and three Sanger
.ab1reads, determine which samples successfully mutated away the BbsI site - RNAi — given a vector and an experimental protocol, design an RNAi plasmid targeting a gene of interest with a given antisense sequence
- CAR-T — given the amino-acid sequence of an anti-CD19 CAR (FMC63 scFv–CD28–CD3ζ), design a pLVX-CAR-IRES-ZsGreen1 expression vector
And More
- Circular plasmid map view alongside the sequence editor
- DNA / RNA / protein modes — single-strand editors for RNA and protein sequences
- Wrap and continuous view modes — multi-line wrapping or a single horizontally scrollable line
- BLAST search — right-click any selection to submit it to NCBI BLAST
- SnapGene history — browse snapshots embedded in .dna files
- Plugin system — extensible architecture for adding custom tools
- Multi-project tabs — switch between plasmids in the sidebar
- Full undo/redo — sequence edits and feature changes
- GenBank / SnapGene / FASTA I/O — read/write .gb/.gbk and protein .gpt; read .dna/.rna/.prot (SnapGene), FASTA and Sanger .ab1
Plugins
LibreGene supports plugins — the dotplot viewer and gel simulator in src/plugins/ are examples. Want to build one? Check the plugin development guide and open a PR. Contributions welcome!
Installation
macOS (Homebrew, Apple Silicon):
brew install --cask dl-li/libregene/libregene
The app is not notarized yet, so on first launch use right-click → Open (or approve it in System Settings → Privacy & Security).
Alternatively, download installers for macOS (dmg), Windows (msi / nsis) and Linux (Flatpak) directly from GitHub Releases.
Quick Start
npm install
npx tauri dev
Requires Node.js ≥ 20, Rust ≥ 1.75, and Tauri v2 prerequisites.
License
GNU General Public License v3.0 — see LICENSE.
Copyright (C) 2025 dl-li
Acknowledgements
Thanks to our contributors liu1568075 and Masterchiefm.
This project also builds on many great open-source tools, including pLannotate, pydna, DNA Chisel, Biopython and RibossFold.
LibreGene MCP questions
How does the MCP integration work?
LibreGene embeds an MCP server on localhost (127.0.0.1:8766) using Bearer-token authentication, allowing LLM agents to operate the plasmid editor while the UI updates live.
What file formats does LibreGene support?
It supports GenBank, SnapGene (.dna, .rna, .prot), FASTA, and Sanger (.ab1) formats.