accept
FreeAccept HQ membership invites from a magic link or raw token.
The English summary is being prepared. Showing the original description.
Tell every open Claude Code desktop session to run /handoff, then collect each session's /resumework command. Triggers: "handoff all sessions", "wrap up every session".
The English summary is being prepared. Showing the original description.
Run score Not tested
See how to installReview the instructions the agent will receive before you install.
Broadcast /handoff to every open desktop session and return one HQ resume
command (/resumework <thread-id>) per session. Never return Claude-native
resume commands (claude --resume, session links) as the resume path.
Arguments: $ARGUMENTS — optional --dry-run (stop after step 2),
--running-only (skip idle sessions), --include <id,...>, --exclude <id,...>.
ccd_session_mgmt tools. In the terminal CLI
or headless (CLAUDE_HEADLESS=1), STOP and say this skill needs the desktop app.Call list_sessions with limit: 100, include_archived: false. The current
session is excluded automatically. Apply --running-only, --include,
--exclude. Skip unattended sessions (scheduled-task runs); send_message
cannot deliver to them.
Show a table: title, running/idle, last activity, session id. Sending a
message starts a turn in each target session, so this is an outward action.
Ask for an explicit go with AskUserQuestion (options: send to all listed /
running only / cancel). --dry-run stops here.
For each target, call send_message with this body (verbatim, fill nothing):
Run
/handoffnow to close out this session. In the handoff, record every active /conduct lane, background agent, PR watch, and other tracked item with its id, state, and next step, plus whether this session was in conduct mode (engine and child model). When it finishes, reply with exactly one final line in the formHQ-RESUME: /resumework <thread-id>. If this session produced no resumable state, replyHQ-RESUME: none (<reason>).
Record the returned delivery status (delivered / queued / error) and
message_id per session. Do not retry errors blindly; report them.
Poll each session with list_events (back off, roughly every 60s, cap ~20
minutes) for the HQ-RESUME: line. As a fallback, match new thread files:
ls -t workspace/threads/T-*.json | grep -v changeset created after the
broadcast, using each thread's session/title fields to map it to a session.
Write workspace/handoff-all/<batch-id>.json where <batch-id> is
HA-YYYYMMDD-HHMMSS. This file is the only input /resume-all trusts.
{
"batch_id": "HA-20261003-031500",
"created_at": "<ISO-8601>",
"resumed_at": null,
"sessions": [
{"session_id": "local_...", "title": "...", "cwd": "...",
"status": "handed_off|no_state|failed|no_reply",
"thread_id": "T-...", "resume_command": "/resumework T-...",
"conduct": {"enabled": true, "engine": "claude", "child_model": "..."}}
]
}
Final message lists one row per session:
| Session | Status | Resume command |
|---|---|---|
| title | handed off / no state / failed / still running | /resumework T-... |
Flag sessions that never answered or errored. Do not run /handoff for this
session unless the user asks.
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Accept HQ membership invites from a magic link or raw token.
The English summary is being prepared. Showing the original description.
Capture qualifying architecture decisions as ADRs in the right HQ or repo location.
The English summary is being prepared. Showing the original description.
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.
The English summary is being prepared. Showing the original description.
Find shallow modules and propose deepening refactors, ranked with file refs; never edits code. Use when code is hard to change or planning a refactor.
The English summary is being prepared. Showing the original description.
Generate dark ASCII banner graphics for HQ and social launch assets.
The English summary is being prepared. Showing the original description.
Compare approaches to best-in-class prior art, merge existing work, verify cruft removal, and recommend the most durable design before PRD work.
The English summary is being prepared. Showing the original description.