[RELEASE] Fish Audio is the 24th chat channel (carries #5055)

[vivekchand]

Publishes #5055 (merged) to PyPI, closing #5045.

Why this needs to ship, not just merge. The channel lives in clawmetry/sync.py, clawmetry/gateway_tap.py, clawmetry/entitlements.py and routes/channels.py — all in the wheel. Ingest runs on the node, so no Fish Audio conversation appears for anyone until the package is on PyPI.

What rides this: #5055 only.

What it adds

fish-audio becomes the 24th chat channel: transcripts ingest from ~/.openclaw/fish-audio/*.jsonl, live events arrive over the gateway tap as fishaudio, and /api/channel/* serves the tab.

People delegate to agents through chat, so a channel ClawMetry cannot see is a conversation that simply does not exist in the product — the shape of the original report, "I message Diya on Telegram and ClawMetry shows nothing".

A chat channel is four independent lists

List	What it drives	Symptom when missing
sync._CHANNEL_DIRS	daemon walks ~/.openclaw/<channel>/*.jsonl	history never ingests
gateway_tap.CHANNEL_NAMES	live WebSocket subscription	live messages never arrive; history works
entitlements.ALL_CHANNELS / CHANNEL_LABELS	the catalogue UI and entitlement read	invisible even with rows in the store
routes/channels.py	the /api/channel/* endpoint	the tab cannot be served

Each omission is silent and looks like a different bug. The second row is the nastiest: a reporter says "old messages show up but nothing new", which points at ingest rather than at a missing subscription.

That rule is now a check

tests/test_channel_four_lists.py asserts every catalogued channel is in both ingest lists, that neither carries an orphan, that each has a label, and that the three lists are the same size. Proven red against a deliberately half-wired channel and green on the tree.

Comparison is on a normalised form, because the disk name and the wire name legitimately differ (fish-audio vs fishaudio) — deriving one list from the other is exactly what the architecture forbids.

Also corrected a comment that would have caused the next half-wired channel: the note above _CHANNEL_DIRS said adding a directory needed "no further wiring required" (false), and still said "21 adapter routes" when there are 24.

Verification

tests/test_entitlement_channel_catalog.py (27 tests) and the count guard pass, reading 24 across every surface.

No-PRD: publishes an already-merged channel adapter; the architecture is documented on the Runtime and Session Observability blueprint.

---

🤖 Generated with Claude Code

https://claude.ai/code/session_01Xb6A5G74JiMe3zHFs1JZEP

[8090-software-factory]

✅ Drift Bot (ClawMetry): no drift detected

Drift Bot analyzed the changed files against this project's blueprints and requirements and found no drift.