labtab

Command-line interface for LabTab - the lab management platform for Principal Investigators. Manage grants, papers, literature, reagents, posters, talks, results, meetings, tasks, projects, and people from your terminal.

Built by Wetware Ltd.

Install

pip install labtab

Requires Python 3.10+.

Shell completion

labtab ships tab completion for bash, zsh, fish, and PowerShell:

# Install completion for your current shell (then restart the shell)
labtab --install-completion

# Or print the completion script to wire up manually
labtab --show-completion

Quickstart

# Pair your terminal with your LabTab account (opens a browser window)
labtab auth login

# Later logins reuse the API key in ~/.config/labtab/config.toml; pair again only when needed
labtab auth login --force

# Pair with a read-only key instead - mutating commands will be rejected
labtab auth login --read-only

# Open the keyboard-first interactive workspace
labtab --interactive

# See which labs you belong to
labtab labs list
labtab labs members list
labtab labs members invite new.member@example.com --role PhD
labtab labs members set-role 42 "PhD"
labtab labs members archive 42
labtab labs members transfer-pi 42
labtab labs settings show
labtab labs settings update --timezone Europe/London

# Review or update your account profile
labtab account show
labtab account update --first-name Alex --last-name Harston

# Manage API keys for scripts and integrations
labtab api-keys list
labtab api-keys create "CI deploy" --scope read:papers --scope write:notes
labtab api-keys revoke 42

# Open a command-line overview of your lab (--limit rows per section)
labtab home
labtab home --limit 10

# Create a meeting - --datetime takes ISO-8601 or a relative form
labtab meetings create --title "Standup" --datetime "tomorrow 10:00" --lab smith-neuro-lab
labtab meetings create --title "1:1" --datetime "next monday 09:00" --lab smith-neuro-lab

# Search across your accessible workspace
labtab search "atlas figures"

# Review and clear unread notifications
labtab notifications list
labtab notifications mark-all-read

# Review PI-only lab activity
labtab audit-logs list
labtab audit-logs show 42

# Create and revoke external share links
labtab shares list --type paper --object-id 12
labtab shares create paper 12 collaborator@example.com --permission comment
labtab shares revoke 7

# Review upcoming lab deadlines
labtab deadlines list

# Track recommendation letters and drafts
labtab letters list
labtab letters show 12
labtab letters drafts 12
labtab letters save-draft 12 --file draft.html
labtab letters status 12 Submitted

# Browse lab wiki pages
labtab wiki list
labtab wiki tree

# Browse and export collaborative notes
labtab notes list meeting 42
labtab notes create meeting 42 --text "Decision captured"
labtab notes update 123 --file follow-up-notes.txt
labtab notes export meeting 42 --format md --output meeting-notes.md

# List your grants
labtab grants list

# Inspect a single grant
labtab grants show gr_abc123

# Show the next 7 days of meetings
labtab meetings list --days 7

# Get JSON for scripting
labtab papers list --json | jq '.[] | .title'

# Curate relevant literature and journal-club collections
labtab literature list --reading-status to_read
labtab literature create --title "Spatial transcriptomics benchmark" \
  --author "Maya Singh" --journal "Nature Methods" --year 2026 \
  --reading-status to_read
labtab literature collection-add 3 12
labtab literature comment 12 --body "Discuss in journal club"

# Track a conference poster
labtab posters create --title "Spatial transcriptomics atlas" \
  --conference ISSCR --presented-date 2026-06-12 --author-id 42 \
  --file poster.pdf

# Manage the reagent catalogue, including plasmid sequence files
labtab reagents list --kind plasmid
labtab reagents create --kind plasmid --name "pLenti-Cas9" \
  --identifier Addgene-52962 --location "Freezer 2 / B4" \
  --sequence-file pLenti-Cas9.dna

# Prepare and archive lab talks
labtab talks list --upcoming
labtab talks create --title "Organoid atlas update" --type Seminar \
  --date 2026-07-01 --speaker-id 42 --project-id 3
labtab talks archive 11

# Capture research outputs and link them to papers, posters, talks, or grants
labtab results create --title "Figure 2 quantification" \
  --description "Validated cell counts" --author 42 --project 3 \
  --file counts.csv --kind data
labtab results link 8 --target-type paper --target-id 22 --relationship supports

# Export your whole lab as a zip
labtab export --format zip --lab smith-neuro-lab

Import Markdown from Obsidian or a folder

Choose the folder whose contents you want to share. The default command uploads its visible .md files for validation and previews new Wiki pages without saving them in LabTab. Add --apply to create them after validation:

labtab notes import-folder ~/Documents/MyVault/LabImports
labtab notes import-folder ~/Documents/MyVault/LabImports --apply

# Select particular files instead of the whole folder
labtab notes import-folder ~/Documents/MyVault --file "Protocols/Cell culture.md" --apply

# Append one file to an existing meeting's collaborative notes
labtab notes import-folder ~/Documents/MyVault --file "Meetings/Planning.md" --meeting <meeting-id> --apply

Use --lab <id-or-slug> to choose the destination lab. Normal CLI write access is required. The read-only Obsidian mirror credential cannot import files.

Original files are never changed. Repeating an import of the same file path with the same account and lab does not create another note. Later source edits remain independent of LabTab. Moving the source file or importing from a different computer gives it a new identity, so select only files you intend to import.

Hidden files, symlinks and LabTab-managed mirror files are skipped. Unsupported content blocks the selection before any import starts. Obsidian [[links]] remain text, and linked files and attachments are not uploaded. Each run supports up to 500 files of 512 KiB each. If an import fails after earlier files succeeded, those imports stay saved. Retry the same command to continue safely.

Interactive workspace

labtab --interactive opens a full-screen workspace over the same API client used by the regular commands:

• ↑ / ↓ move through resources and records; Enter opens a record in LabTab
• / filters the current view as you type
• Ctrl+K searches across your LabTab workspace
• Ctrl+G sends a natural-language goal to LabTab AI and reviews its proposals
• x confirms a quick action or AI proposal; d rejects an AI proposal
• b focuses the resource list, r refreshes, and q exits

The selected record shows its equivalent labtab ... --json command. Agents and scripts should continue using those non-interactive commands: their JSON output and exit behavior are unchanged. AI actions are never applied without a confirmation, and destructive proposals receive explicit destructive approval.

Commands

Commands that fetch structured data support --json for machine-readable output. Most lab-scoped commands accept --lab <slug> to override the default lab from your config; PI-only audit logs accept the lab ID used by the API permission check.

Configuration

labtab reads and writes its config at ~/.config/labtab/config.toml. Override the path with the LABTAB_CONFIG_FILE environment variable.

[api]
base_url = "https://labtab.app/api/v1"

# Optional when the browser app uses a different origin from the API.
[web]
base_url = "https://labtab.app"

[auth]
api_key = "lab_api_xxxxxxxx"
user_email = "pi@example.com"
default_lab_id = 42
default_lab_slug = "smith-neuro-lab"

The config file is written with 0600 permissions so other users on the machine cannot read your API key.

If you later move the product domain, you can rewrite just the hostname while preserving the rest of the API path:

labtab auth set-base-url --host labtab.io

Pydantic models

The package ships two sets of models:

• labtab.models (re-exports labtab.models.api) - hand-curated, lenient models the CLI uses internally. They tolerate extra or renamed backend fields so a CLI install doesn't break when the server ships new columns.
• labtab.models.generated - a strict, auto-generated snapshot of the backend's OpenAPI schema. Useful if you want to script against the API directly with precise types.

Regenerating the strict snapshot

# with a running backend at localhost:8000
datamodel-codegen \
  --url http://localhost:8000/api/v1/schema/ \
  --output src/labtab/models/generated.py \
  --output-model-type pydantic_v2.BaseModel \
  --input-file-type openapi \
  --target-python-version 3.10

# or from a snapshotted schema
cd backend && python manage.py spectacular --file /tmp/schema.yml
datamodel-codegen --input /tmp/schema.yml \
  --output src/labtab/models/generated.py \
  --output-model-type pydantic_v2.BaseModel \
  --input-file-type openapi \
  --target-python-version 3.10

License

MIT. See LICENSE.