switchboard
Skill4 starsChanged 3 months agoMITView on GitHub
Use when the user asks to open the switchboard dashboard, see every running Claude Code session, check plan usage or weekly token budget, or review their Claude Code usage patterns. Starts a local dashboard on http://localhost:7823 showing status, token burn and messages for every session on this Mac.
5 other projects use a version of this.
Adopted by
Most recently taken up by
- tanishq286/tq-forgetheir own version, first seen Nov 15, 2026
- shayaansultan/switchboardtheir own version, first seen Oct 15, 2026
- keez97/switchboardtheir own version, first seen Oct 11, 2026
- TentacleOpera/switchboardtheir own version, first seen Oct 8, 2026
- Wondermonger-daydreaming/claude-skills-librarytheir own version, first seen May 30, 2026
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- TentacleOpera/switchboard★ 215same name
- tanishq286/tq-forge★ 8same name
- Wondermonger-daydreaming/claude-skills-library★ 7same name
- shayaansultan/switchboard★ 4same name
- keez97/switchboard★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repositoryOne other file it gives its agents.
- CLAUDE.mdCLAUDE.md
The file
We split this skill into the rules it gives an agent
5 rules. 0 are common, 4 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 5 edited versions we read keep these.
4
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 15
Start the server if it isn't already up, then open the page.
- Very specific: line 18
If up, skip to step 3. Nothing else to do — one server serves every session.
- Distinctive: line 19
Start it detached, and read back the URL it printed (the port walks upward if 7823 is taken):
- Distinctive: line 24
If the log shows an error instead of a URL, show it to the user and stop — do not retry.
- Distinctive: line 27
Do not run the server in the foreground and do not poll it.
“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
switchboard
A local dashboard for every Claude Code session running on this Mac. macOS + Node 18+.
$1 selects what to do. Default (no argument) is open.
open
Start the server if it isn't already up, then open the page.
- Already running?
curl -sf -m 2 http://localhost:7823/api > /dev/null && echo up- If
up, skip to step 3. Nothing else to do — one server serves every session.
- If
- Start it detached, and read back the URL it printed (the port walks upward if 7823 is taken):
If the log shows an error instead of a URL, show it to the user and stop — do not retry.nohup node "${CLAUDE_PLUGIN_ROOT}/claude-sessions.js" > /tmp/switchboard.log 2>&1 & sleep 3; cat /tmp/switchboard.log open <url>and tell the user the URL in one line.
Do not run the server in the foreground and do not poll it. It is a long-lived process the user leaves running; it is not something to wait on.
report
node "${CLAUDE_PLUGIN_ROOT}/claude-sessions.js" --report
Prints the derived weekly budget — what 100% is in tokens, what is left, the run rate, and whether it runs dry before the reset. Show the output as-is; it is already formatted.
Read the caveat out of the README before interpreting it: this is raw token volume, and cache reads dominate it. It is a consistent yardstick week over week, not a number Anthropic would recognise. Don't present it as an official figure.
patterns
node "${CLAUDE_PLUGIN_ROOT}/claude-sessions.js" --patterns [days]
Prompts by hour and weekday, most-used tools, models, effort, prompt lengths, interrupts and
concurrent sessions. [days] defaults to 7. No prompt text is stored or shown.
Notes
- The dashboard shows the text of prompts. It binds to loopback only. If the user asks to expose it on the network, tell them what that publishes (prompts, working directories, session names, unauthenticated) before doing it.
- Plan-usage bars read the OAuth token from the login Keychain, so macOS prompts on first run. Denying it is fine — everything else still works.
- State lives in
~/.switchboard/. Deleting it loses day-by-day history, nothing else.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

