Klau CLI

Import job data from CSV or Excel files into Klau and get optimized dispatch plans — no code required.

Built for ops teams, dispatchers, and billing clerks who export daily work orders from their existing system and need them loaded into Klau for dispatch optimization.

Install

dotnet tool install -g Klau.Cli

Requires .NET 9.0 runtime or later.

Get Started

Step 1: Log in

klau login

This will prompt for your Klau email and password, then automatically create and store an API key. You only need to do this once — the key is saved to ~/.config/klau/credentials.json and used for all future commands.

If you already have an API key from Settings > Developer in the Klau dashboard:

klau login --api-key kl_live_your_key_here

Step 2: Preview your data

Before importing, preview how your file will be mapped. This sends nothing to Klau — it just reads your file and shows you what it found:

klau import daily-orders.csv --dry-run

or with an Excel file:

klau import work-orders.xlsx --dry-run

You'll see which columns were mapped, any data quality warnings, and a preview of the first 5 rows. Fix any issues in your export before proceeding.

Step 3: Import

klau import daily-orders.csv --date 2026-04-03

The CLI will:

1. Check that your Klau account is set up (drivers, trucks, yards, dump sites)
2. Validate your data (addresses, container sizes, job types)
3. Import jobs into Klau
4. Report results

Step 4: Import + Optimize

klau import daily-orders.csv --date 2026-04-03 --optimize

This imports your jobs and then runs Klau's dispatch optimizer. You'll get back a plan grade, flow score, and assignment summary.

To also export the optimized dispatch plan as a CSV:

klau import daily-orders.csv --date 2026-04-03 --optimize --export dispatch-plan.csv

Supported File Formats

Format	Extensions	Notes
CSV	.csv, .tsv, .txt	Auto-detects delimiter (comma, tab, semicolon, pipe)
Excel	.xlsx, .xls	Reads the first worksheet, first row as headers

The CLI handles quoted fields, BOM markers, and mixed line endings automatically.

Column Mapping

The CLI automatically maps your column headers to Klau's job fields. It recognizes header names from common hauler dispatch systems — no configuration needed for most exports.

How it works

1. Exact match — Customer Name → CustomerName
2. Fuzzy match — Cust, Service Name, Billing Name → CustomerName
3. Prefix match — C_ADDR1, C_CITY, C_STATE → SiteAddress, SiteCity, SiteState

Recognized headers

Klau Field	Your column might be called
CustomerName	Customer Name, Cust, Account Name, Service Name, Billing Name, Company
SiteAddress	Address, Street, Service Address, C_ADDR1, FullAddy, Full Address
SiteCity	City, Site City, Service City, C_CITY
SiteState	State, ST, Service State, C_STATE
SiteZip	Zip, Zip Code, Site Zip, C_ZIP
JobType	Service, Type, Job Type, Service Type, Service Code, Service Description
ContainerSize	Container, Size, Container Size, Size Value, Yard, Yards
ExternalId	Order #, Order Nbr, Order Number, Work Order, WO, PO, Reference, Ticket
RequestedDate	Date, Service Date, Requested Date, Delivery Date, Scheduled Date
Notes	Notes, Instructions, Comments, Special Instructions, Billing Notes
Priority	Priority, Order Priority
TimeWindow	Time Window, Delivery Window

Saving and reusing mappings

On first import, the CLI saves the inferred mapping to .klau-mapping.json next to your file. Future imports from that directory reuse the saved mapping automatically — no re-inference needed.

To use a specific mapping file:

klau import orders.csv --mapping /shared/our-column-mapping.json

Data Validation

The CLI validates your data before sending it to Klau and gives you clear guidance when something needs fixing.

Pre-flight account check

Before importing, the CLI verifies your Klau account is set up for dispatch:

  Pre-flight readiness check:
    ✗ Drivers: not configured
    → Add drivers: Klau dashboard > Drivers, or via SDK
    ✗ Trucks: not configured
    → Add trucks: Klau dashboard > Trucks, or via SDK
    ✓ Account ready (85% configured)

Row-level validation

The CLI catches data quality issues before they hit the API:

• Missing addresses — "Row 5: no address for 'Acme Corp' — geocoding will fail, job may not be routable"
• Invalid container sizes — "3 rows have non-standard container size ('99') — expected: 10, 15, 20, 30, 35, 40"
• Unmapped job types — "8 rows have unmapped job type ('RO-DUMP & RETURN3L') — configure service code mappings in Settings > Company"
• Duplicate order numbers — "2 duplicate external IDs — duplicates will be rejected by the API"

If more than half your rows are missing addresses, the CLI flags it as a likely column mapping error and suggests re-running with --dry-run.

Automated Import (Watch Mode)

For daily automation — e.g., your billing system drops a CSV export to a shared folder every morning — use watch mode:

klau import watch --folder /exports --pattern "*.csv" --optimize

What watch mode does

• Monitors the folder for new CSV/XLSX files
• Waits for each file to finish writing before processing
• Runs the full import + validation + optimize flow
• Moves successful files to processed/
• Moves failed files to failed/ with a .error file explaining what went wrong
• Writes dispatch plans to output/ (when --optimize is set)

Designed for reliability

Watch mode is built to run unattended for months:

• Lock file — Prevents two watchers on the same folder
• Heartbeat — Writes .klau-heartbeat every 60 seconds (monitor for staleness)
• Error recovery — A bad file doesn't crash the watcher; it's moved to failed/ and processing continues
• Retention — Processed files are cleaned up after 30 days (configurable with --retain-days)
• Graceful shutdown — Responds to Ctrl+C and SIGTERM cleanly

Watch mode options

Option	Description	Default
--folder	Folder to watch (required)	—
--pattern	File glob pattern	*.*
--date	Dispatch date, or today for dynamic	today
--optimize	Run optimization after each import	false
--retain-days	Days to keep processed files	30

Authentication

Key resolution

The CLI checks for your API key in this order:

1. --api-key flag (highest priority)
2. KLAU_API_KEY environment variable
3. ~/.config/klau/credentials.json (stored by klau login)

Commands

klau login                          # Interactive: email/password → auto-create API key
klau login --api-key kl_live_...    # Store an existing key
klau status                         # Show current auth state
klau logout                         # Remove stored credentials

For non-admin users

The interactive klau login flow creates an API key automatically, which requires admin access in Klau. If you're not an admin, ask your admin to generate a key for you in Settings > Developer, then use:

klau login --api-key kl_live_the_key_they_gave_you

Standalone Optimization

Already imported jobs through the dashboard or SDK? Run optimization separately:

klau optimize --date 2026-04-03

Optimization modes

Mode	Description
full-day	Re-optimize the entire day (default)
new-job	Fit newly added jobs into the existing plan
rebalance	Rebalance assignments across drivers

klau optimize --date 2026-04-03 --mode new-job --export dispatch-plan.csv

Diagnostics

klau doctor

Checks your environment, authentication, API connectivity, and account configuration in one command. Shows exactly what's working and what needs attention.

Multi-Tenant Support

For companies with multiple divisions or subsidiaries, each with their own fleet:

klau tenant list                    # List available tenants
klau tenant set <tenant-id>         # Set default tenant for all commands
klau import orders.csv --tenant <id>  # One-off tenant override

AI Agent Integration (MCP)

The CLI can run as an MCP server for AI agents like Claude Desktop, Cursor, and other MCP-compatible tools:

klau mcp

This exposes import, optimize, doctor, and status as MCP tools over stdio. AI agents can call them directly with structured parameters and get JSON responses — no shell parsing needed.

MCP client configuration

Add to your MCP client config (e.g., Claude Desktop claude_desktop_config.json):

{
  "mcpServers": {
    "klau": {
      "command": "klau",
      "args": ["mcp", "--api-key", "kl_live_your_key_here"]
    }
  }
}

If you've already run klau login, you can omit --api-key — the MCP server uses stored credentials.

JSON Output

All commands support structured JSON output for scripting and CI/CD:

klau import orders.csv --output json
klau optimize --date 2026-04-03 --output json
klau doctor --output json

Every command emits a consistent envelope:

{
  "command": "import",
  "status": "success",
  "exitCode": 0,
  "data": { "imported": 45, "skipped": 2 },
  "error": null
}

Non-interactive mode

For CI/CD pipelines and automation scripts, use --yes to skip all interactive prompts:

klau import orders.csv --date 2026-04-03 --optimize --yes

Shell Completions

Enable tab completion for your shell:

# Bash
echo 'eval "$(klau completion bash)"' >> ~/.bashrc

# Zsh
echo 'eval "$(klau completion zsh)"' >> ~/.zshrc

# Fish
klau completion fish | source

# PowerShell
klau completion powershell >> $PROFILE

Troubleshooting with --verbose

When something isn't working, --verbose shows HTTP request details on stderr:

klau doctor --verbose

  [verbose] GET /api/v1/companies/me -> 200 (805ms)
  [verbose] GET /api/v1/companies/go-live-readiness -> 200 (534ms)

Share this output with support to diagnose connectivity or API issues.

Updating

klau upgrade

The CLI checks for updates in the background and will notify you when a new version is available.

All Commands

Command	Description
klau login	Authenticate and store credentials
klau logout	Remove stored credentials
klau status	Show auth state and configuration
klau doctor	Diagnose environment, auth, and account readiness
klau import <file>	Import a CSV or XLSX file
klau import watch	Watch a folder for files to import automatically
klau optimize	Run dispatch optimization for a date
klau tenant list	List available tenants
klau tenant set	Set the default tenant
klau mcp	Run as an MCP server for AI agents
klau completion <shell>	Generate shell completion scripts
klau upgrade	Update to the latest version

Global options

Option	Description
--api-key	Override stored API key
--tenant	Override default tenant for this command
--output json	JSON output mode
--yes / -y	Skip interactive prompts
--no-color	Suppress colored output (also via NO_COLOR env var)
--verbose / -v	Show HTTP request details on stderr

Import options

Option	Description	Default
--date	Dispatch date (YYYY-MM-DD)	today
--mapping	Path to column mapping file	auto-detected
--optimize	Run dispatch optimization	false
--export	Export dispatch plan CSV	—
--dry-run	Preview mapping and validate without importing	false

Optimize options

Option	Description	Default
--date	Dispatch date (YYYY-MM-DD)	today
--mode	Optimization mode: full-day, new-job, rebalance	full-day
--export	Export dispatch plan CSV	—

Example: Daily Operations Workflow

Morning — one command:

klau import todays-orders.xlsx --date 2026-04-03 --optimize --export dispatch-plan.csv

Result:

  Reading todays-orders.xlsx... 47 rows (XLSX)

  Column mapping:
    Customer Name   →  CustomerName
    Service Address →  SiteAddress
    Service City    →  SiteCity
    Service State   →  SiteState
    Site Zip        →  SiteZip
    Size Value      →  ContainerSize
    Order Nbr       →  ExternalId
    Service Date    →  RequestedDate

  Preview (5 of 47 rows):
    Customer                Address                 City          Type      ExternalId
    ----------------------  ----------------------  -----------   --------  ----------
    Acme Construction       123 Industrial Way      Springfield   DELIVERY  WO-1001
    Metro Demolition        456 Oak Ave             Shelbyville   PICKUP    WO-1002
    Harbor Freight          789 Commerce Dr         Capital City  SWAP      WO-1003

  Pre-flight readiness check:
    ✓ Account ready (100% configured)

  Importing 47 jobs for 2026-04-03...
    ✓ Imported: 45  |  Skipped: 2
    ✓ Auto-created: 3 customers, 5 sites
    ! Row 12: invalid container size "ABC"

  Optimizing dispatch...
    ✓ Grade: A (92/100)  |  Flow: 87/100
    ✓ Assigned: 43/45  |  Unassigned: 2

  Exported to dispatch-plan.csv

  Done in 14.2s

Overnight automation:

klau import watch --folder /exports --optimize

The billing system drops a file at 4 AM. By 5 AM, the dispatch plan is waiting in output/.

Troubleshooting

"No API key found"

Run klau login to authenticate, or set the KLAU_API_KEY environment variable.

"Blocking issues must be resolved before import"

Your Klau account is missing required configuration (drivers, trucks, yards, or dump sites). Set these up in the Klau dashboard before importing jobs.

"Over half the rows are missing addresses"

The address column in your file likely isn't being mapped correctly. Run klau import <file> --dry-run to see the column mapping and fix it. You can create a custom .klau-mapping.json to override the auto-detection.

"Unmapped job type"

Your file uses service codes that Klau doesn't recognize (e.g., "RO-DUMP & RETURN3L"). Configure service code mappings in Settings > Company in the Klau dashboard to map your codes to Klau job types (DELIVERY, PICKUP, DUMP_RETURN, SWAP).

Column mapping is wrong

Delete the .klau-mapping.json file next to your CSV and re-run with --dry-run. The CLI will re-infer the mapping. If auto-detection doesn't work for your format, create a manual mapping file:

{
  "Your Customer Column": "CustomerName",
  "Your Address Column": "SiteAddress",
  "Your City Column": "SiteCity",
  "Your State Column": "SiteState",
  "Your Zip Column": "SiteZip",
  "Your Service Column": "JobType",
  "Your Container Column": "ContainerSize",
  "Your Order # Column": "ExternalId"
}

License

MIT