Why I built GlanceFlow: knowing what Claude Code is doing, at a glance
GlanceFlow puts one calm, plain-English checklist above the Claude Code prompt: the plan, live progress, and a clear signal when Claude needs you.
Watching Claude Code work usually means scrolling past a wall of tool calls, diffs and command output. For engineers, half of it is noise. For everyone else, it's unreadable. I kept asking the same question: Claude is working… but on what? How far along is it? Is it waiting for me?
GlanceFlow is my answer. It's a Claude Code mod that puts one calm, plain-English checklist above the prompt. Before Claude uses a tool, it lays out its plan in 2 to 8 short steps. Each step fills in as Claude works, with a quiet line on what it's doing right now and, when a step is done, one sentence on what it got done. The header says about how much time is left, learned from your own history. When Claude needs you, the checklist says so clearly, and tells you where to answer.
It also helps with a few things that go wrong in long sessions:
- Sending a secret by mistake. A message that looks like it holds a password, API key, token or card number is held back. Send it again within two minutes if you really meant it.
- Showing a secret on screen. Emails, phone numbers, keys and card numbers that match its patterns are masked on screen, which helps when you share your screen. Claude still reads the original text. Pattern matching cannot catch every secret.
- Claude going off in the wrong direction. With "Approve the plan first" on, Claude shows its plan and waits for you to press Start.
- A job that quietly stalls. If nothing happens for three minutes, GlanceFlow says Claude may be stuck and how to stop it.
- Losing the thread in a long chat. Before tidying up the chat, GlanceFlow asks Claude to save a checkpoint of the goal, decisions and next step to help it resume after compaction.
Your history stays on your computer. GlanceFlow is free and open source, and the source is on GitHub: github.com/Antreas-Strb/glanceflow
I built it for people who use Claude Code every day, engineers or not, and want to look away, come back, and know where things stand in one glance.
— Antreas
Continue reading
All posts
Launch story: Roomcomm
Codex and other agents from different owners share one room. Every owner reads the same transcript. A room is a link: hand it to your agent, someone hands it to theirs, and they talk.

OpenMatter Network and HOL Release Proposed Standards for Verifiable AI Compliance and Agentic Security
OpenMatter Network and HOL released proposed standards for Zero-Knowledge Boundary Compliance for Autonomous Agents. Public comment is open through October 31, 2026.

Claude Code tried to dump your Kubernetes secret
Claude Code ran kubectl get secret db-credentials -o yaml, which exports the whole Secret. HOL Guard paused it: Allow just this once or Keep blocked.
