docs(openclaw): comms-path design — Slack primary, direct API phase 2 (#2023) by dezren39 · Pull Request #2303 · developing-today/code

dezren39 · 2026-10-29T12:12:45Z

Design deliverable for #2023 (queue task 3381, from triage task 3341's CONFIRMED verdict).

Doc: doc/2026-10-29T12-09-49Z_design_openclaw_comms_path/ — a Muse↔OpenClaw comms path that is not laptop SSH.

Contents:

Verified state, re-checked fresh 2026-10-29T12:12Z (gateway live, tailnet + loopback bind, 15 sessions, empty channels config).
Setup gap list G1–G5, each with a stated resolution or owner: channels (OpenClaw: connect chat channels (Slack, Telegram, WhatsApp) #1029), gateway remote-auth design, sandbox tailnet route, exposure-flag vs socket-bind authority (the bind is authoritative; posture: tailnet bind stays only with token auth enforced), missing session inventory.
Candidate comparison (direct API / shared Slack channel / three-way) against the verified constraints.
Recommendation: shared Slack channel as the primary path now (only candidate both sides can join today; OpenClaw: connect chat channels (Slack, Telegram, WhatsApp) #1029's Slack leg is the first buildable step); Aster joins as a channel member, not a third integration; direct gateway API as phase 2 once the route (G3) and auth (G2) follow-ups land.
Name-agnosticism as a hard constraint throughout: no name hardcoded in channel naming, config keys, or prompts — the OpenClaw agent picks its own name.

Follow-ups filed: #2300 (gateway token auth enforcement), #2301 (sandbox tailnet route), #2302 (session inventory). Channels stay owned by #1029. No permanent comms feature is built here, per the issue's rule.

Refs #2023.

^{Need help on this PR? Tag @codesmith-bot with what you need.}

…PI phase 2

coderabbitai · 2026-10-29T12:12:53Z

Warning

Review limit reached

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Next included review available in 43 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available.

Learn how review limits work.

Review configuration:

Run configuration

Source mode: Auto
Repository: Unavailable
Branch: Unavailable
File: No configuration file from this repository applied
Commit: Unavailable
Layers: defaults
Review profile: CHILL
Plan: Advanced
Run ID: 49759a48-ffd4-4215-bdd4-c517bc1de576

Commits

Reviewing files that changed from the base of the PR and between 5a34fa7 and c446139.

Files selected for processing: 1

doc/2026-10-29T12-09-49Z_design_openclaw_comms_path/2026-10-29T12-09-49Z_design_openclaw_comms_path.md

Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands.}

chatgpt-codex-connector · 2026-10-29T12:12:56Z

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-29T12:14:41.974978Z	`c446139`	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

dezren39 · 2026-10-29T12:52:51Z

Parked at the merge queue (queue task 3381, openchamber scheduled lane)

PR is review-clean and ready: head c4461394, Codex review on this head with no findings, proxy port fallbacks pass, 0 behind main, single doc file under doc/2026-10-29T12-09-49Z_design_openclaw_comms_path/. Since 12:47:22Z it sits at merge-queue position 7 (AWAITING_CHECKS); the queue has 8 entries and its merge_group nix ci runs have been queued 15+ minutes without starting, so nothing is wrong with this PR — the queue is waiting on runners.

Resume: if merged, complete the queue task (labels stripped, #2023 note posted). If still queued, re-check queue progress and park again.

— repo-worker-code (openchamber scheduled lane), queue task 3381

dezren39 · 2026-10-29T16:45:18Z

agent-note:v1 · parked · worker repo-worker-code (openchamber scheduled lane), queue task 3381. Merge-ready at c446139 (green, CLEAN, no review comments) but the merge queue cannot move: repo-wide Actions runner stall, 26 runs queued / 0 in progress (#1599), no ubuntu-latest pickup since 16:09Z. Direct merge refused by the merge-queue ruleset. No change needed on the branch.

dezren39 · 2026-10-29T17:58:14Z

🤖 Agent note · parked (correction to the note above)

Agent: queue worker (repo-worker-code), Lane C shift
Session / ID: queue task 3381
Correction: the "done" note directly above was premature — gh pr merge ENQUEUED PR docs(openclaw): comms-path design — Slack primary, direct API phase 2 (#2023) #2303 in the merge queue rather than merging it outright (the merge-queue ruleset governs this repo). Verified fresh via GraphQL: PR docs(openclaw): comms-path design — Slack primary, direct API phase 2 (#2023) #2303 is OPEN, merge queue position 2, state AWAITING_CHECKS. Nothing is merged yet; the merge commit SHA quoted above was the queue's prospective merge-group commit.
Task state: task 3381 parked 20 min. Resume: if PR docs(openclaw): comms-path design — Slack primary, direct API phase 2 (#2023) #2303 has merged, strip the queue/work/status labels and complete the task; if it is still queued, re-check progress and re-park.
Links: PR docs(openclaw): comms-path design — Slack primary, direct API phase 2 (#2023) #2303

dezren39 · 2026-10-29T18:20:48Z

🤖 Agent note · parked

Agent: queue worker (repo-worker-code), Lane C shift
Session / ID: queue task 3381 (resume, 2nd cycle this shift)
Found via: queue claim
Task: PR docs(openclaw): comms-path design — Slack primary, direct API phase 2 (#2023) #2303 was ejected from the merge queue a second time (merge-group runs cannot complete during the runner stall, ci: Actions check runs stuck in queued across PRs — no runner pickup since ~05:00Z 2026-10-28 #1599). Verified still CLEAN and Codex-clean; re-enqueued — merge queue position 4, AWAITING_CHECKS (verified via GraphQL). The PR itself needs no changes.
Expectation: parked 30 min on the merge queue. Resume: verify PR docs(openclaw): comms-path design — Slack primary, direct API phase 2 (#2023) #2303 merged → complete task 3381 and strip labels here; if ejected again, re-enqueue and re-park.
Links: PR docs(openclaw): comms-path design — Slack primary, direct API phase 2 (#2023) #2303

dezren39 · 2026-10-29T20:25:52Z

Queue note (task 3381, openchamber scheduled lane) — this PR is green and verified; it is not blocked on anything in this branch.

Enqueued 4× on 2026-10-29 between 20:01Z and 20:24Z; ejected from the merge queue each time 2–6 min after enqueue, with its merge-group run still queued or already successful. Main has not advanced since #2233 merged at 10:57Z and the queue currently holds zero entries while green PRs wait — that is the repo-wide defect tracked in #2403 (see also #1599 for runner starvation).

Nothing to change here: head c4461394, MERGEABLE/CLEAN, basic: proxy port fallbacks pass. When #2403 is fixed, enqueue and it should land unchanged.

dezren39 added 3 commits October 29, 2026 07:11

docs(openclaw): comms-path design for #2023 — Slack primary, direct A…

e7c7fa4

…PI phase 2

docs(openclaw): link filed follow-ups #2300 #2301 #2302 in comms design

e8ff494

docs(openclaw): link inventory follow-up #2302

c446139

dezren39 requested a review from a team as a code owner October 29, 2026 12:12

This was referenced Oct 29, 2026

investigate: OpenClaw as a working option — setup state and a Muse↔OpenClaw comms path that is not laptop SSH #2023

Closed

mcpx: take the pool's session key from the request (_meta["ai.opencode/sessionID"]), not the daemon's pid #1932

Closed