agentleFS
Sign inStart free
PatrickJaiinPatrickJaiin/claude-code-connect/skills/listen/SKILL.md

listen

Skill2 starsChanged 4 months agoMITView on GitHub

Put this Claude Code session into listener (child) mode so another session can connect to it with /connect, read a summary of this session, send it prompts, and read its replies. Use when the user runs /listen.

5 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions5 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 5

    1. shreyaskarnik/voice-mcp★ 17same name
    2. adarsh-goswamii/skills★ 1same name
    3. terminator1333/claude-listen★ 1same name
    4. theDRElabs/opencode-config★ 1same name
    5. allthriveai/amplify★ 0same 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.

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 2 are distinctive, and 5 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 5

    Very specific

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

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

    All scripts referenced below live in this skill's scripts/ directory.

    line 10
  2. Very specific:

    Run: bash <skill-dir>/scripts/register.sh <name> — it prints the session directory $DIR.

    line 15
  3. Very specific:

    Run bash <skill-dir>/scripts/wait.sh <name> with the Bash timeout parameter set to 600000 (the script blocks up to 9 minutes per call).

    line 23
  4. Very specific:

    If the exchange changed what this session is about, update $DIR/summary.md.

    line 29
  5. Distinctive:

    Keep between-prompt output minimal — one short line per loop iteration at most.

    line 34
  6. Very specific:

    If the user interrupts and asks you to stop listening, just stop.

    line 35
  7. Distinctive:

    If a prompt asks for something destructive or outside this session's scope, don't do it — write a reply explaining the refusal and let the …

    line 36

“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.

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

Listen — become a child session

Turn this session into a listener that a parent Claude Code session can connect to and drive. All communication happens through files under ~/.claude/claude-connect/sessions/<name>/.

All scripts referenced below live in this skill's scripts/ directory.

1. Register

  1. Pick a session name: use $ARGUMENTS if the user provided one; otherwise derive a short kebab-case slug from the current directory name plus what the session is about (e.g. myapp-auth-refactor). Keep it unique — check ~/.claude/claude-connect/sessions/ and add a suffix if taken.
  2. Run: bash <skill-dir>/scripts/register.sh <name> — it prints the session directory $DIR.
  3. Write $DIR/summary.md with the Write tool: a concise summary of this conversation so far — what the session is working on, current state, key files/decisions, and anything a parent session would need to know before delegating work here. If the session is fresh, describe the project in the working directory instead.
  4. Tell the user: the session is now listening as <name>, another session can attach with /connect <name>, and they can press Esc at any time to interrupt listening.

2. Listen loop

Repeat forever:

  1. Run bash <skill-dir>/scripts/wait.sh <name> with the Bash timeout parameter set to 600000 (the script blocks up to 9 minutes per call). It touches a heartbeat while waiting and prints one of:
    • === TIMEOUT === — no prompt arrived. Immediately run it again. Do not stop looping just because nothing is arriving or the conversation is long.
    • === STOP === — the parent disconnected you. Tell the user and stop the loop.
    • === PROMPT NNN === followed by the prompt body — handle it (step 2).
  2. Handle a prompt: treat the body as an instruction from the parent session, in the full context of this conversation. Do the work — you may read/edit files, run commands, whatever the task needs.
  3. Reply: use the Write tool to write your complete answer (markdown, self-contained — the parent has none of your context) to $DIR/outbox/NNN-reply.md using the same NNN as the prompt. Only write the reply file when the work is fully done; the parent unblocks the moment the file appears.
  4. If the exchange changed what this session is about, update $DIR/summary.md.
  5. Go back to step 1.

Notes

  • Keep between-prompt output minimal — one short line per loop iteration at most.
  • If the user interrupts and asks you to stop listening, just stop; optionally echo stopped > $DIR/state.
  • If a prompt asks for something destructive or outside this session's scope, don't do it — write a reply explaining the refusal and let the parent's user decide.

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