agents-view
Reviewed
at commit 095e52e
New version, not yet reviewed
A pane listing this session's subagents (Agent tool and workflow agents) with their status; select one to see its prompt, text, tool calls and results.
0 upvotes
Log in to upvote or save- Made by
- ushironoko
- Pinned commit
-
095e52e5dd74a855a62724037f3b12a21234e767 - What we checked
-
Opens a pane listing this session's subagents and workflow agents with status, time and token counts. Selecting one shows its prompt, text and tool calls, read from the session's transcript files. Selecting an agent also starts a separate Haiku request on your own Claude account that writes a short Japanese summary of it, which uses tokens.
Review details
Static review of the source at the pinned commit; not run. Full re-review of the new version (the diff was too large for a delta review). Hooks: session.start (registers /subagents and a 500 ms redraw loop), command.run, agent.spawn, turn.start, turn.step and tool.call for Workflow (observe only), ui.render for the pane. It reads, with $.fs, the current session's transcript and subagent files under CLAUDE_CONFIG_DIR or ~/.claude/projects/<project>/<session id> (it lists the projects folder only to find the folder holding this session id), and uses $.process.run for head, tail, perl and grep over the session's own transcript to read large files or extract workflow names. When the user selects an agent, it runs 'claude -p --model claude-haiku-5-5 --tools "" --disable-slash-commands --strict-mcp-config --no-session-persistence' with a system prompt and passes the agent's label, task prompt (3000 characters) and last 40 tool or text entries on stdin; the child process gets AGENTS_VIEW_CHILD=1 so the mod does nothing inside it, and a 120 s timeout applies. Output is clipped and stripped of control characters before drawing. It writes no files, makes no network calls itself, and keeps state in memory. Not reviewed: the claude CLI.
- What it uses
-
- Reads your prompts: no
- Sees tool calls: yes
- Blocks or changes tool calls: no
- Reads or writes files: yes
- Runs programs on your machine: yes
- Calls the model (uses tokens): yes
- Sends data over the network: no
- Stores data: no
- Shows something on screen: yes
Made this mod? Add a badge to your README
Paste this into your README. The badge says reviewed once the current version has been reviewed, and links readers to this page.
[](https://modsforclaude.com/mods/agents-view/)