{"slug": "claude-agent-map-see-every-agent-and-what-its-doing", "title": "claude-agent-map: see every agent, and what it’s doing", "summary": "Developer iha2 released claude-agent-map, a read-only Claude Desktop plugin that opens an \"Agent map\" pane showing the main session and every subagent spawned beneath it, with each agent's type, description, superlite task ids, status color (green running, red aborted or blocked, grey done), tier, elapsed time and tokens. The pane is opened with the `/claude-agent-map` command, by asking Claude, or by replying `approve map` at superlite's plan approval, and installing superlite from the marketplace pulls the plugin in as a dependency. The tool addresses parallel agent runs that fail quietly in the UI, turning a swarm of agent IDs into a view of who spawned whom and which tasks remain open.", "body_md": "[Field Notes](https://contextfirst.blog/tag/field-notes/)\n\n# claude-agent-map: see every agent, and what it’s doing\n\nA short note on claude-agent-map: a Claude Desktop pane that shows the main session and every subagent — tier, time, tokens, and tasks — so parallel agent runs stay readable.\n\n[Superlite](https://github.com/iha2/superlite?ref=contextfirst.blog) will happily put a dozen agents to work at once. That is the point. It is also the problem.\n\nWhen the build is going well, the transcript is a waterfall of launches, reviews, and checkpoints. When it is not, you need to know which agent stalled, which task it owned, and whether the coordinator is still thinking or just waiting. Grepping `progress.md` and `metrics.jsonl` works. It does not feel like watching the run.\n\nSo I split out [claude-agent-map](https://github.com/iha2/claude-agent-map?ref=contextfirst.blog): a small Claude Desktop plugin that opens a read-only **Agent map** pane beside the transcript.\n\n## What it does\n\nIt shows the main session at the top, and every subagent under the agent that spawned it. For each one you get:\n\n- **Who it is** — type and short description\n- **What it is working on** — superlite task ids when those records are present (✓ once checkpointed)\n- **How it is doing** — green while running, red if aborted or blocked, grey when done\n- **What it cost so far** — tier, elapsed time and tokens, plus a detail view for model, parent, ledger lines, and last message\n\nThe pane only reads. It never steers a run.\n\nOpen it with `/claude-agent-map`, optionally pointing at a project (or its `.superlite` folder). You can also ask Claude to open it, or reply `approve map` when superlite offers the pane at plan approval. Installing superlite from the marketplace pulls this plugin in as a dependency.\n\n## Why it’s useful\n\nParallel agent workflows fail quietly in the UI long before they fail in the tests. Eighteen agents is not “visibility.” It is a pile of IDs. The map turns that pile into a picture: who spawned whom, which tasks are still open, and which nodes went red.\n\nJust enough structure that you can leave a swarm running and still answer “what is happening right now?” without opening six files.\n\nThe repo is here: [https://github.com/iha2/claude-agent-map](https://github.com/iha2/claude-agent-map?ref=contextfirst.blog).", "url": "https://wpnews.pro/news/claude-agent-map-see-every-agent-and-what-its-doing", "canonical_source": "https://contextfirst.blog/claude-agent-map-see-every-agent/", "published_at": "2026-10-27 02:32:42+00:00", "updated_at": "2026-10-27 02:41:17.551510+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "developer-tools", "ai-products"], "entities": ["claude-agent-map", "Claude Desktop", "Superlite", "iha2", "GitHub"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/claude-agent-map-see-every-agent-and-what-its-doing", "markdown": "https://wpnews.pro/news/claude-agent-map-see-every-agent-and-what-its-doing.md", "text": "https://wpnews.pro/news/claude-agent-map-see-every-agent-and-what-its-doing.txt", "jsonld": "https://wpnews.pro/news/claude-agent-map-see-every-agent-and-what-its-doing.jsonld"}}