[28/36] Add OC-140 WebSocket lifecycle task

[cjohnsto-nz]

Summary

Adds the supervisor planning task for first-class WebSocket connection lifecycle UX, including editor, CodeLens, status bar, and Copilot tool expectations.

Stack

Base: feature/oc-130-protocol-layout-stability (PR #59)
Head: supervisor/add-websocket-lifecycle-task
Immediate child: feature/oc-140-websocket-lifecycle-ux (PR #61)

Validation

• Documentation links and task references checked
• npm run compile
• npm test
• npm run build
• Packaged missio-0.8.0.vsix
• Installed missio.missio@0.8.0 locally

[APKiwi]

Verdict: approve with nits. The task doc is internally consistent (goal, acceptance, tests, out-of-scope all align) and its acceptance criteria are a good yardstick, in fact two of them (secret redaction, no leaked listeners) are exactly where the implementation PR falls short, so the doc is stronger than the code that implements it. One nit: AGENT_PROGRESS says OC-130 accepted while README calls OC-130 a follow-up task, both edited in this PR, pick one.

[cjohnsto-nz]

Reviewed this as the planning-only PR for OC-140.

The task document is internally consistent and is the acceptance contract for the implementation. The secret-redaction and listener-cleanup observations concern implementation PR #61; they are not missing implementation from this planning branch, and I am assessing them on PR #61.

The OC-130 status mismatch is valid on this intermediate branch, but it is already fixed later in the stack by PR #66, commit 390b991 (Register OC-150 in agent wiki). That commit changes the compatibility summary to: “OC-000 through OC-140 are implemented...” Later planning commits continue progressing the same status line for OC-160 and OC-170.

I tested a duplicate correction on PR #60 before pushing it. The immediate edge PR #60 -> PR #61 remained clean, but the stack-tip simulations PR #60 -> PR #71 and PR #60 -> PR #72 both produced a conflict in docs/open-collection-gap-analysis/README.md. I therefore removed the unpushed duplicate instead of knowingly breaking the stack. The exact disposition is FIXED_LATER in PR #66, not an unresolved suggestion.

I also corrected the malformed base/head placeholders in this PR description. The task links resolve, PR #60 and immediate child PR #61 are live MERGEABLE/CLEAN, and this PR's build and security checks pass.