feat(web): keep the PR hub current and expose missing review actions

[rynfar]

The pull-request hub loses its remembered filters and useful details while reopening or refreshing a PR, and it lacks controls for pending workflow approval and reverting a merge.

Port previously approved upstream #9188, #9334, and #9294: remember list controls, keep cached details visible while refreshing, preserve current summary state, and add supported GitHub workflow-approval and revert actions with server-side permission checks. Keep the header usable at narrow widths and the existing Pylon branding and host support. Scope cached details to their environment and tolerate browsers that deny storage access.

Web and the desktop renderer share these controls, including remote environments. Mobile has no PR hub; its shared contracts still typecheck. GitLab, Azure DevOps, and Bitbucket retain their supported action sets. The unrelated desktop build and chat-scrolling fixes now have separate PRs #281 and #282.

Validation: 612 focused tests across 11 server, web, and contract files pass; server, web, mobile, desktop, contracts, and client-runtime typechecks pass; targeted lint passes with two existing default-prop warnings; formatting and diff checks pass. Integrated browser verification passes using an isolated Pylon environment with providers disabled. At 980×960, header controls fit without horizontal overflow. Selecting Oldest shown, leaving for a thread, and returning retains the sort. Closing/reopening the detail panel and refreshing it preserve visible cached details; a DOM observer saw no loading-status insertion during those operations. No real GitHub workflow-approval or revert operation was executed; those action and permission branches are covered by focused tests. The baseline Base UI nativeButton warning was recorded; the candidate pass added no console error.

Before / after at 980×960:

Before	After

Video: close, reopen, and refresh a cached PR panel (29 seconds; unused recording area cropped).

hub-after.mp4

Implemented with GPT-6 Astra through the Codex harness.

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	13.6 KiB	13.6 KiB	−28 B (−0.2%)	15.1 KiB	✅
Codex	Thread snapshot wire	7.1 KiB	7.1 KiB	−10 B (−0.1%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.5 KiB	6.5 KiB	−18 B (−0.3%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	56.5 KiB	56.5 KiB	0 B (0.0%)	66.4 KiB	✅
Codex	Live turn messages	8	8	0 (0.0%)	21	✅
Claude	Total thread wire	13.6 KiB	13.6 KiB	−27 B (−0.2%)	15.1 KiB	✅
Claude	Thread snapshot wire	7.1 KiB	7.1 KiB	0 B (0.0%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.6 KiB	6.5 KiB	−27 B (−0.4%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	57.3 KiB	57.3 KiB	0 B (0.0%)	66.4 KiB	✅
Claude	Live turn messages	8	8	0 (0.0%)	21	✅

Baseline: aa12838 · PR result: 7b171f7 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 111.2 KiB
• Claude decoded thread snapshot: 111.9 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.