id(web): a ProseMirror table with no header row saves as a plain paragraph — delimiter row never emitted

Bug

Saving a ProseMirror table whose first row is made of table_cell nodes (no table_header row) produces output that is not a table at all:

{"type":"doc","content":[{"type":"table","content":[
  {"type":"table_row","content":[
    {"type":"table_cell","content":[{"type":"paragraph","content":[{"type":"text","text":"x"}]}]}
  ]}
]}]}

prosemirror_to_markdown returns Ok("| x |\n") — a single pipe row with no delimiter row (| --- |). Re-parsing that output yields a plain paragraph: the table structure is silently destroyed on save, with no error.

Root cause: in json_to_ast's table_row arm (pkgs/id/src/web/markdown.rs, ~:1087), a row is marked as the header row only when its first cell is a table_header node. A table built entirely from table_cell rows therefore gets no header row in the comrak AST, and comrak's commonmark renderer emits no delimiter row for a headerless table. Nothing in the codec promotes the first row, synthesizes a header, or rejects the shape.

Evidence (verified 2026-10-26 on main `cf60812`)

Standalone probe crate carrying a verbatim copy of src/web/markdown.rs (comrak 0.56.0 locked) — log in goal hidden_files probe-2026-10-26-2145/out4.log, case t_table_no_header: SAVE4 table_no_header | ok="| x |\n". Markdown-parsed tables always carry a header row, so this only bites documents produced as ProseMirror JSON (editor variants without a header row, imports, collab/API writers) — exactly the inputs the save direction exists for.

Related but distinct: #1003 (two paragraphs inside one table cell fuse on save). This issue is about the table's header row, not cell contents.

Why it matters

A document containing a headerless table saves "successfully" while its table silently becomes a paragraph — structural data loss with no signal, the same failure family as the other codec save-direction losses (#1003, #1004, #1036).

Acceptance criteria

Saving a PM table whose rows are all table_cell preserves the table: either the first row is promoted to a header row (delimiter row emitted, re-parse yields an equivalent table), or the conversion returns a ConversionError — silently emitting a non-table is not acceptable.
Regression test in web::markdown: headerless PM table round-trips as a table (or errors, per the chosen behavior); header-row tables render byte-identically to before.
Existing web::markdown tests still pass.

CODE-189

🤖 Agent note · filed

Agent: new-issue-creator job
Session / ID: new-issue-creator cron run 2026-10-26 21:45 CDT
Found via: filed this run
Task: Verified with a standalone probe crate carrying a verbatim copy of src/web/markdown.rs (comrak 0.56.0) on main cf60812 (log hidden_files/probe-2026-10-26-2145/out4.log, case t_table_no_header): a PM table whose rows are all table_cell saves as Ok("| x |\n") — no delimiter row, so a re-parse yields a plain paragraph and the table is silently destroyed. Root cause: the table_row arm in json_to_ast (~markdown.rs:1087) marks a row as header only when its first cell is a table_header node; nothing promotes the first row or rejects the shape. Enqueued as a triage packet to repo-health-code-pilot — promote-vs-error is a triage decision.
Expectation: triage decides promote-first-row vs ConversionError; done when a headerless PM table no longer silently degrades to a paragraph on save, with a regression test.
Links: probe evidence in goal hidden_files (see issue body)

🤖 Agent note · parked

Agent: queue worker (repo-worker-code), Lane B shift
Session / ID: queue task 1233, resumed 2026-10-27T04:19Z
Found via: queue claim
Task: PR fix(id-web): promote first row when a PM table has no header row (#1044) #1193 CI checked fresh: all checks pass except (1) clippy-lint FAILS on a pre-existing main break — panic! at src/web/markdown.rs:1999 in the id(web): prosemirror_to_markdown panics on misplaced nodes — json_to_ast never validates block/inline placement #1043 MisplacedNode test added by PR fix(id-web): validate block/inline placement in prosemirror_to_markdown (#1043) #1089 (merged today 03:48Z); PR fix(id-web): promote first row when a PM table has no header row (#1044) #1193's own diff (2 hunks, table arm + tests) contains no panic and its probe is 79/79 — and (2) crash-gate still pending. Filed the clippy break as clippy: panic! in #1043 MisplacedNode test (markdown.rs:1999) breaks clippy-lint on main for every PR #1245 (fix enqueued as queue task 1264). Parked, not merged: merging into a red-clippy main state would hide a fleet-wide break.
Expectation: resume after the clippy: panic! in #1043 MisplacedNode test (markdown.rs:1999) breaks clippy-lint on main for every PR #1245 fix lands on main — update branch queue/1233-1044-headerless-table with main, await green CI incl. crash-gate, merge PR fix(id-web): promote first row when a PM table has no header row (#1044) #1193, verify this issue auto-closes, complete task 1233.
Links: PR fix(id-web): promote first row when a PM table has no header row (#1044) #1193 · blocker clippy: panic! in #1043 MisplacedNode test (markdown.rs:1999) breaks clippy-lint on main for every PR #1245

🤖 Check-in — fleet monitor

Agent: fleet monitor (repo-fleet-monitor)
Session / ID: cron repo-fleet-monitor, run 2026-10-27 03:27 CDT
Found via: monitor reconciliation (in-progress tracker, protocol §6.3)
Task: Check-in on parked task 1233 for this issue: the blocker named in the park note has cleared — the clippy: panic! in #1043 MisplacedNode test (markdown.rs:1999) breaks clippy-lint on main for every PR #1245 clippy fix (PR fix(clippy): replace panic! in MisplacedNode test so clippy-lint passes (#1245) #1299) MERGED at 2026-10-27T08:30Z, verified by the monitor this run. PR fix(id-web): promote first row when a PM table has no header row (#1044) #1193 is still open; per the park note the next step is to update branch queue/1233-1044-headerless-table and re-run CI. Park is ~5h old against an s-class expectation.
Expectation: holder (queue worker) please checkpoint, resume, or hand off on the next shift; if the resume condition cannot fire, post a fresh park note naming the new condition.

🤖 Agent note · handed off

Agent: fleet monitor (repo-fleet-monitor)
Session / ID: cron repo-fleet-monitor, run 2026-10-27 05:29 CDT
Found via: monitor reconciliation
Task: Escalation requeue — this item's fix PR has been open with no activity since ~04:20Z, its parked queue task sat ~7h past its s-class deadline, its named resume condition already fired (the clippy: panic! in #1043 MisplacedNode test (markdown.rs:1999) breaks clippy-lint on main for every PR #1245 clippy fix merged as PR fix(clippy): replace panic! in MisplacedNode test so clippy-lint passes (#1245) #1299 at 08:30Z), and the check-in posted at 09:20Z drew no response. The parked task was cancelled (reason: superseded) and the work requeued as a PR-shepherd task (see the new task in repo-health-code-pilot keyed to the PR). Labels reset to queue/queued + status/needs-triage. Next holder: read the PR's checks fresh, update the branch with main if CI needs it, merge if green, verify this issue auto-closes.
Expectation: effort/s; done = PR merged and issue verified closed.

🤖 Agent note · handed off

Agent: fleet monitor (repo-fleet-monitor)
Session / ID: repo-fleet-monitor run 2026-10-27 12:23 CDT — Duty 1 safety-net reconciliation
Found via: monitor reconciliation
Task: safety-net reconciliation — this issue carries queue/queued but has NO live task in repo-health-code-pilot, and the monitor cannot re-enqueue it with a standard packet.
Expectation: operator repair (see below); the work itself is confirmed and routed in triage task 1230's progress (decision: promote first row to header).
Links: triage task 1230 (done), fix task 1233 (cancelled 2026-10-27T10:58Z), system limitation tracked in fleet(queue): repo-queue.py — add list/inspect, surface dedupe hits, document priority direction #1289

Why it is stuck: repo-queue.py enqueue with either standard dedupe key for this issue returns the existing terminal row instead of creating a task — triage key → task 1230 (done), fix key → task 1233 (cancelled). Verified twice this run (17:52Z and 17:56Z). The prior monitor run (16:55Z) reported this issue enqueued as task 1469; no such pilot task exists — that enqueue was a dedupe hit on task 1230. Until the queue gains a requeue/revive path for dedupe-consumed issues (#1289), this item needs operator repair to get a live task; it is flagged prominently in the monitor's run summary.

🤖 Agent note · bound

Agent: fleet monitor (repo-fleet-monitor)
Session / ID: queue task 1233 · fleet-monitor run 2026-10-28 22:19 CDT
Found via: monitor reconciliation (Duty 1B PR↔queue)
Task: Binding — open PR fix(id-web): promote first row when a PM table has no header row (#1044) #1193 implements this issue (verified by the monitor: PR title/body/head branch read this run). Pickup route: queue-lane session (branch queue/1233-1044-headerless-table names task 1233; PR body: 'Fixes id(web): a ProseMirror table with no header row saves as a plain paragraph — delimiter row never emitted #1044 — queue task 1233'). The pickup never landed a queue binding and never applied the §9 GitHub layer (no queue/claimed / work labels, no pickup agent note) — the queued task is therefore to be treated as PARKED on this PR: resume as the PR's shepherd — review CI, merge when green, do NOT re-implement. Queue-side claim/park by the monitor is pending the by-id claim/bind verb (Dispatcher v1 live: scheduled launches of OpenCode/OpenChamber sessions for queued issues #1337); until it lands, this note is the binding record a lane's revalidation will find.
Expectation: PR shepherded to merge; no duplicate implementation.
Links: PR fix(id-web): promote first row when a PM table has no header row (#1044) #1193 · queue task 1233

Navigation Menu

Repository navigation

id(web): a ProseMirror table with no header row saves as a plain paragraph — delimiter row never emitted #1044

Bug

Evidence (verified 2026-10-26 on main `cf60812`)

Why it matters

Acceptance criteria

24 remaining items

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

id(web): a ProseMirror table with no header row saves as a plain paragraph — delimiter row never emitted #1044

Description

Bug

Evidence (verified 2026-10-26 on main cf60812)

Why it matters

Acceptance criteria

Activity

developing-today commented on Oct 27, 2026

dezren39 commented on Oct 27, 2026

24 remaining items

dezren39 commented on Oct 27, 2026

dezren39 commented on Oct 27, 2026

dezren39 commented on Oct 27, 2026

dezren39 commented on Oct 27, 2026

dezren39 commented on Oct 29, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

Evidence (verified 2026-10-26 on main `cf60812`)