Command

PePersonal JEJust Electrolyte RTRelease Timer TtTacTile CSClearSource AtAthletic

TacTile›Inbox — on deck·49 in this wave

Wave

Inbox — on deck

Briefs Nova has teed up and are ready to run — each task's prompt is a complete, self-contained brief.

TacTile › Inbox — on deck

39/49 done10 open3 open blockers

Selected taskclear

verify:unit has been red for weeks on one node_modules bug — html-encoding-sniffer@6 is CJS but its dep is ESM-only

DoneOwner · ait_59fe2322

No dependencies in this wave — nothing gates it, and nothing waits on it.

Show the prompt

You are Viv, the agent for the TacTile lane of Command.

Open the lane and work there:

```bash
cd ~/Command/tactile && claude
```

Read that lane's CLAUDE.md and lane.json before you touch anything — they
carry your identity, your scope, and the hazards specific to this lane.
Read state from the hub (https://internal.tactile.to) or ask. Never guess at it.

## The task

**verify:unit has been red for weeks on one node_modules bug — html-encoding-sniffer@6 is CJS but its dep is ESM-only** · done

Where it sits: TacTile › Inbox — on deck

## The brief on file

CLOSED 2026-08-08 — NOT A DEFECT. Do not reopen without running `pnpm install` first.

The claim: `tests/unit/card-bridge-behavior.test.ts` cannot collect because `html-encoding-sniffer` is CJS and `require()`s `@exodus/bytes`, which is ESM-only, so `verify:unit` is permanently red and `bash scripts/verify/gate.sh` can never exit 0.

It is false, and it has now been reported as a hard blocker by THREE separate sessions. Every time, the same commit gates green on a freshly-installed tree:

  $ pnpm install --frozen-lockfile
  $ bash scripts/verify/gate.sh
  → GATE_EXIT=0, 17 passed / 1 failed of 18, verify:unit ✓
    (the one red is verify:secrets-rotation, the allowed-red)

Verified on origin/main at 126d1fcd, at 51405460, and at af099e35.

The ESM story does not hold up either — read the installed manifests rather than reasoning from the package names. Both `html-encoding-sniffer@6.0.0` and `@exodus/bytes@1.15.1` declare `engines.node: "^20.19.0 || ^22.12.0 || >=24.0.0"`, so the declared range is satisfied by the Node in use.

ROOT CAUSE is a stale worktree install: a worktree that has not re-run `pnpm install` since a new workspace package landed fails to transform a suite, and vitest reports that as failing assertions rather than a load error. See the `stale-worktree-install-fakes-test-failures` note.

WHAT IT COST, so the pattern is worth recognising: on 2026-08-08 a session treated this as blocking, abandoned its third work item, and left two finished PRs (#293, #294) unmerged for a failure that did not exist. Both merged clean minutes later — the integration tree of the two together gated at exit 0 on the first try.

If a session reports this again: `pnpm install` in that worktree, re-run, and only then investigate.

## How to work this

- Do the work, don't just plan it. If the fix is obvious and safe, make it.
- Surface anything needing Matthew's sign-off instead of deciding it yourself.
- Short replies. Lead with the answer, then stop.
- Finish by saying what changed, what you verified, and what you'd do next.

To doIn progressDoneBlockedPick a wave on the left · tap a task to expand it

Lanes

Inbox — on deck

verify:unit has been red for weeks on one node_modules bug — html-encoding-sniffer@6 is CJS but its dep is ESM-only