factcircuit
FactCircuit: an auditable claim-and-evidence verification loop with immutable provenance, staged checks, bounded retrieval, and reproducible evaluation.
Register histological brain sections to BrainGlobe atlases using vision-language and image-generation models.
DataPython · JavaGitHub @greenpoloSelf-installOwner unconfirmed
Paste into Claude or ChatGPT for installation instructions suited to your environment.
Have you tried it? One tap is enough. No sign-in needed.
No updates detected yet. GitHub releases will appear here.
Owners of public repositories under a personal GitHub account can sign in to verify ownership and manage the introduction themselves. No administrator approval is needed.
Sign in, generate a verification file, and commit it to this repository’s default branch within 30 minutes. Files on other branches or forks do not count. Then check the file below to manage the project without administrator approval. Keep the file to retain your access.
If needed, you will go to sign-in. After signing in, return here and press your chosen verification button again.
Installable projects can be verified without a deployment domain. Use GitHub sign-in for a personal repository, or the repository verification file above for an organization repository.
Ask about other verification options
Loading the full README. The saved excerpt is shown below.
Agent-driven registration of histological brain sections to BrainGlobe atlases.
LangSlice gives a vision-language agent tools to register a stack of histology
sections to a BrainGlobe atlas. The agent inspects
the anatomy, sets positions and cutting angles, aligns sections in-plane,
and optionally fits ANTs deformations. It chooses regions and settings and
can use an image model to trace borders where useful.Read the full README ↗LangSlice gives a vision-language agent tools to register a stack of histology sections to a BrainGlobe atlas. The agent inspects the anatomy, sets positions and cutting angles, aligns sections in-plane, and optionally fits SyN deformations. It chooses regions and settings and can use an image model to trace borders where useful.
Source images stay unchanged. Work is checkpointed and undoable, with
coordinate and label maps, registration.json, and QuickNII / VisuAlign
exports. Coronal sections of the Allen mouse atlas are the most tested case.
Download and extract the source, open a terminal in the folder with
environment.yml, then:
conda env create -f environment.yml
conda activate langslice
langslice login # ChatGPT sign-in; or set OPENAI_API_KEY / GEMINI_API_KEY (see .env.example)
On Windows, run the same conda env create and conda activate commands in
Miniforge Prompt or PowerShell with conda initialized. Python 3.11 on Windows
is tested in CI. Atlases are downloaded by BrainGlobe on first use into the
user's .brainglobe folder. Python 3.11 or newer is required; the environment
includes the deformable engine (FireANTs on torch), the Claude Desktop server and the ABBA
launcher.
langslice-job, Claude Desktop through MCP, or the built-in agent with
langslice linear run sections/.Tools and options · Tool plugins · File formats · Architecture · Full documentation
Cite the project metadata in CITATION.cff (GitHub's "Cite this
repository" button). BSD-3-Clause; see LICENSE.