Documentation

CLI

qiklog send, tail-file, and key management from the terminal.

The qiklog CLI ships logs from your terminal — useful for scripts, CI, and tailing local files.

Install / run from source

From the repository (requires .NET 9 SDK):

dotnet run --project src/QikLog.Cli -- --help

A packaged single-file binary is planned for Homebrew and Scoop.

Global options

Option	Env var	Default	Description
`--api`	—	`http://localhost:5080`	API base URL
`--key` / `-k`	`QIKLOG_API_KEY`	—	API key when auth is enabled

send — one log line

dotnet run --project src/QikLog.Cli -- send \
  --source demo \
  --message "deploy finished" \
  --level info

With an API key:

dotnet run --project src/QikLog.Cli -- send \
  -s demo -m "deploy finished" \
  --key "$QIKLOG_API_KEY"

Against the hosted API (invite beta):

export QIKLOG_API_KEY='ql_…'   # from Manage or an invite
dotnet run --project src/QikLog.Cli -- send \
  --api https://qiklog-api.up.railway.app \
  -s demo -m "hello from the CLI" -l info

The CLI sends Authorization: Bearer <key> (same credential the API also accepts as X-QikLog-API-Key). There is no interactive OAuth login for send / tail-file — that is intentional for agent-style ingest.

Exit code 0 on 202 Accepted; non-zero on failure with stderr details.

tail-file — ship a local file

Follows a file like tail -f and POSTs each new line:

dotnet run --project src/QikLog.Cli -- tail-file ./app.log --source mybox

Press Ctrl+C to stop. Useful for dev boxes and quick demos without changing application code.

key create — development only

Creates a key via POST /v1/dev/keys (API must be in Development):

dotnet run --project src/QikLog.Cli -- key create --name "local dev"

Prints JSON with the new key. Store it in QIKLOG_API_KEY or your password manager.

Typical workflows

Smoke test after deploy

qiklog send -s production -m "release v1.2.3 live" -l info --api https://api.qiklog.com --key "$QIKLOG_API_KEY"

Watch in browser

Open /tail/production on the web app while shipping logs.

Next steps

You are through the core docs. Revisit Getting started for the mental model, or run Quickstart again on a fresh machine.