blog
Clawdline Blog — orchestration and loops for Claude Code and Codex
What we think about the problem itself: what is actually difficult about running several AI coding sessions at once, which decisions Clawdline made because of it, and what it deliberately does not do.
Claude Code agents fixing bugs you report from your phone, while they refactor
Round two of refactoring Clawdline with Clawdline: four bugs reported from a phone, fixed in parallel, and a review that stopped a fix worse than the bug it fixed.
Refactoring with Claude Code agents: one board, separate reviewers, a measurement per step
How we used Clawdline's Board, built-in roles and review Sessions to refactor Clawdline itself in one day: daemon CPU from 12.6% to 1.3%, 4,383 lines of product code removed.
Clawdline vs Herdr: which fits the sessions you already run?
Compare Clawdline and Herdr by where your Claude Code or Codex sessions run, how you reconnect from another machine or phone, and what happens after an agent finishes.
Coding agents waiting on builds: what one night's callback estimate shows
Clawdline's Heavy Work callback lets a coding agent stop checking a queued build. One night of polling records suggests 74.3% fewer tokens for those waiting turns.
Ten Claude Code sessions, one you open: what a session status lets you do
Every Clawdline session status names one action, so a list of sessions tells you which one to answer, what you still owe, and which tab is safe to close without opening any of them.
Managing Claude Code and Codex together: Clawdline, Happy, Orca and the official remotes
Compare how Clawdline, Happy, Orca, Anthropic Remote Control and Codex on mobile reach your existing sessions or start new work, and where each is limited.
Two Claude Code sessions, one checkout: stopping them committing each other's work
Several agent sessions in one repository share one git index. The pre-commit check that refuses, the half that can only warn, and how Clawdline knows which line of work a path belongs to.
What a scheduled Claude Code task costs, and when a cron job is the right answer
A timer costs nothing; the session it starts does. What a Clawdline schedule buys over a shell script on cron, which of your chores actually needs one, and the measurement behind both answers.
Claude Code context window full? Clawdline's handoff vs just messaging another session
Why Clawdline has a handoff, what the mechanism is, and how it differs from sending a message to another session. The dividing line is whether they already have your context.
Can a Claude Code agent spawn its own agents? Clawdline gives it one level
An agent that hands work out will eventually hand out work you did not ask about. Where Clawdline puts the depth limit so a session cannot argue its way past it, and what the worst case costs.
Does Claude Code run locally or in the cloud? What leaves your machine, and what a stolen login gets
Claude Code runs in your terminal, and Clawdline keeps it there while your phone watches. What the relay carries, what our side can see field by field, and what a key nobody else holds costs you.
What a runaway Claude Code task costs, and how Clawdline lets you cancel at minute three
A task going the wrong way looks exactly like one going well. Clawdline asks every task it dispatches to state its plan before it starts, which is what moves the decision to minute three.
When Claude Code says it's done: delivered, reviewed, or actually on main?
One green tick hides four different facts. Clawdline keeps delivered, reviewed, landed and safe-to-close apart, and will not mark a task landed until git says the commit is on your branch.
How to manage multiple Claude Code sessions: Clawdline attaches to the ones you already started
A free local daemon and web console for macOS and Linux. Clawdline lists Claude Code and Codex sessions already running on your machine, with four separate statuses for each.