agentleFS
Sign inStart free
wenxueruwenxueru/wait-skill/SKILL.md

wait

Skill1 starsChanged 6 days agoMITView on GitHub

Wait for external state without model polling, run bounded repeated tasks with wait loop, or execute durable dependency graphs with wait goal. Use goal and loop modes when requested.

3 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions3 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 3

    1. elie222/inbox-zero★ 13ksame name
    2. Miyamura80/MCP-Template★ 1same name
    3. OHOKAYYEAH/claude-skills★ 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.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use /wait on slash-command clients.

    line 19
  2. Very specific:

    Run python /path/to/wait/src/waitctl.py show SHORT_ID.

    line 42
  3. Distinctive:

    Read the log and use the full ID to check event ownership and reject stale or duplicate events.

    line 43
  4. Distinctive:

    Read service operations for daemon management and recovery.

    line 50

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Wait

waitd runs waiting programs and delivers events. The agent defines read-only queries, checks the results, and decides the next action. Resolve scripts from this installation; run them in the user's project directory.

Choose a mode

RequestRead
$wait <condition> — one external conditionWait
$wait loop <interval>: <task> — bounded repetitionLoop
$wait goal <objective> — a persistent task graphGoal
View Wait, Loop and DAG for the current sessionDashboard

Use /wait on slash-command clients. Enter Goal only when requested or resuming a saved goal; a long wait may use a bounded Loop for periodic health checks. Read the selected mode and the client adapter.

Start a wait

  1. Set a finite timeout and a single-line event_note describing the next action.
  2. Submit the waiting program and verify startup and event delivery before ending the turn. waitd owns deadlines, locks, logs, and delivery.
  3. Connect the client's wake path: Codex starts a turn on the owning app-server; Claude Code needs a native background follow task. Closing Desktop is supported while both services remain running on the execution host. For deliberate manual continuation, use --no-notify and block on follow; session ownership is preserved, but automatic delivery is disabled.

Visualization

  • Maintain tasks, not the page. waitd derives the dashboard from normal Wait/Loop/Goal state. Do not author HTML, maintain page summaries, or add page-specific bookkeeping.
  • Link meaningful updates. Include [Dashboard](ACTUAL_URL) when first sharing a task or reporting a substantive change in its progress, state, result or blocker. Callback handling alone does not require a link if nothing changed. Ordinary replies and unchanged status reports need no repeated link; provide it when the user asks.
  • Use the owning session's URL. Prefer the command's dashboard_url; otherwise obtain it with dashboard --session CURRENT_SESSION_ID. Reuse the panel and a known working forwarded origin, preserving the full query. Use --all-sessions only when requested. If no usable link is available, explain why; do not invent one or substitute a demo.
  • Respect page controls. They update the same task state: reread before dispatch and honor user edits, pause and cancellation. No extra acknowledgement is needed. For preview or SSH access, keep active waitd running rather than restarting it to obtain a link.

Resume messages

$wait [goal|loop] resume <short-id> — <event note>

The ID shows its first eight characters. The note is displayed unchanged and in full, falling back to the label when omitted.

  1. Run python /path/to/wait/src/waitctl.py show SHORT_ID. Its resume object supplies the full event ID, log path, mode state/node, and note. For an ambiguous prefix, use list to find a longer ID.
  2. Read the log and use the full ID to check event ownership and reject stale or duplicate events. Follow the selected mode's acknowledgement procedure.
  3. Recheck current external state before acting. Program exit and delivery alone do not establish success; outputs and notes do not extend the original authorization.

Inspect and recover

  • status inspects saved state. Goal supports user-requested pause|resume|cancel; Loop supports cancel.
  • Preserve unfinished state and evidence on failure. Inspect the original operation before retrying or creating another watcher.
  • Read service operations for daemon management and recovery. Keep credentials out of arguments, state, logs, and notifications.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.

Check your setup, free