feat(viz): add viz monitor web dashboard

Introduces the `humanize monitor web` command and its Flask + SPA
dashboard for RLCR sessions, plus the accumulated review-phase
fixes and UX polish.

Feature
- Per-project session discovery and summary (state.md, goal tracker,
  round summaries, review results, methodology reports).
- Pipeline canvas with snake-path node layout, SVG connectors, pan /
  zoom, and per-round flyout details.
- Session-detail page carries a below-canvas live-log panel with a
  three-state toggle (collapsed / normal / expanded) driven by
  --log-h and a single grid-template-rows declaration. Clicking a
  pipeline node auto-collapses the log; closing the flyout restores
  the prior state.
- Incremental pipeline update (_updatePipelineIncremental) appends
  or replaces individual nodes and redraws only the SVG connectors,
  preserving zoom / pan across round additions.
- Diff-based home refresh: only cards whose rendered HTML changed
  are replaced; unchanged cards stay untouched.
- Per-session SSE live-log streaming from ~/.cache/humanize with
  snapshot / append / resync / eof / retained-replay protocol.
- Active-log selection mirrors the CLI monitor: pick the top
  round's codex-run log, with codex-review / gemini-run /
  gemini-review fallbacks.
- REST + WebSocket + SSE transports. Localhost binds skip auth;
  remote binds require a bearer token (Authorization header for
  fetch, ?token= query for SSE) plus origin-matched CSRF gating on
  every mutating route.
- Ops-menu Preview Issue flow: calls local Claude CLI to generate
  a sanitized methodology report (humanize issue taxonomy), shows
  a preview modal, and submits via `gh issue create --repo
  humania-org/humanize`. Falls back to a copy-to-clipboard path
  when gh is unauthenticated.
- Active-session dynamic progress visualization, methodology report
  generation via local Claude CLI, and a comprehensive frontend test
  suite.

Correctness and safety
- broadcast_message uses set.difference_update to avoid the
  UnboundLocalError from Python treating `-=` as a local rebind.
- Session path validation rejects `..`, dotfiles, symlink escapes;
  plan-file reads are constrained to the project / session dir.
- Remote-mode auth fail-closed when no token is configured.
- CSRF layer enforced on every mutating route, including wildcard
  binds and IPv6 literals.
- Parser accepts table-form acceptance criteria and legacy /
  single-letter criterion ids; multi-id table cells are split
  correctly; COMPLETE verdict only fires on terminal lines.
- Finalize-phase classification scoped to the live round, and
  skip-impl sessions correctly classify round 0 as review-only.

Transport and watcher
- watcher.py starts a cache-log observer per active session that
  broadcasts `round_added` when a new round-N-*.log file appears,
  so the dashboard switches the live-log pane to the new round
  without waiting for the next state.md write.
- SessionWatcher.on_session_created hook primes cache-log observers
  for brand-new sessions; every event retries the hook so the
  state-dir-before-cache-dir race eventually heals.
- Non-destructive toggleTheme (no DOM rebuild). The analytics route
  is the only surface whose charts bake CSS vars into SVG at render
  time, so that route alone re-renders on theme change.
- Remote-mode slow polling (10s) uses the same targeted refresh
  helpers as the WS path so card counters and pipeline nodes catch
  up without a full-page reload.
- Session-detail refresh race guard: after awaiting
  /api/sessions/<id>, re-check the active route + layout skeleton
  before mutating DOM.

Install and CLI
- Preflight codex_hooks and idempotent hook merge when paths contain
  spaces.
- Relaxed runtime detection and tolerate unwritable default shim
  paths.
- `humanize monitor web` foreground mode passes --host explicitly,
  IPv6 brackets in the printed URL, readiness probe is fail-closed.

Parser and API
- /api/sessions summary carries every field the terminal
  `humanize monitor rlcr` status bar needs (push_every_round,
  ask_codex_question, agent_teams, drift, stall count,
  ultimate_goal, tasks_active / deferred, active_log_path,
  git_status, etc.).
- cache_logs entries carry round / tool / role / path / basename
  for deterministic client-side selection.

Version: 1.17.0.

Signed-off-by: Chao Liu <chao.liu.zevorn@gmail.com>

Author: zevorn

.gitignore

@@ -5,6 +5,9 @@ temp
 /.claude/settings.json
 /.claude/scheduled_tasks.lock
 
+# Local Codex CLI marker (empty file occasionally left behind in worktree)
+/.codex
+
 # Humanize state directories (runtime-generated, project-local)
 .humanize/
 .claude-flow/

README.md

@@ -68,8 +68,15 @@ Requires [codex CLI](https://github.com/openai/codex) for review. See the full [
    humanize monitor skill      # All skill invocations (codex + gemini)
    humanize monitor codex      # Codex invocations only
    humanize monitor gemini     # Gemini invocations only
+   humanize monitor web        # Browser dashboard for the current project
    ```
 
+   The `humanize monitor web` subcommand launches a per-project browser dashboard
+   that layers on top of the same data sources the terminal monitors read. It runs
+   in the foreground by default; pass `--daemon` for the background tmux launcher
+   and `--host` / `--port` / `--auth-token` to configure remote access. See the
+   upgrade note: `/humanize:viz` has been removed in favour of `humanize monitor web`.
+
 ## Monitor Dashboard
 
 <p align="center">

docs/drafts/viz-monitor-web.md

@@ -0,0 +1,107 @@
+# Draft: Optimize viz-dashboard — Merge into `humanize monitor` as a Web View
+
+## Background
+
+The `feat/viz-dashboard` branch currently introduces a `/humanize:viz` Claude
+slash command and a local visualization dashboard for Humanize. While the
+dashboard does show some data, the visualization of a *live, dynamically
+running RLCR loop* is not clear enough today: status, progress per round, and
+streamed log output are hard to follow as a loop progresses.
+
+Separately, Humanize already ships a CLI-side monitoring capability that the
+user runs in another terminal (NOT inside Claude Code):
+
+```bash
+source <path/to/humanize>/scripts/humanize.sh   # or add to .bashrc / .zshrc
+
+humanize monitor rlcr        # RLCR loop
+humanize monitor skill       # All skill invocations (codex + gemini)
+humanize monitor codex       # Codex invocations only
+humanize monitor gemini      # Gemini invocations only
+```
+
+This monitor capability already captures live state (RLCR rounds, skill / Codex
+/ Gemini invocations, log output). The web dashboard does not need to invent
+its own capture pipeline — it should consume what `humanize monitor` already
+provides.
+
+## Goal
+
+Optimize the viz-dashboard branch so that:
+
+1. The dashboard becomes a **web view** layered on top of the existing
+   `humanize monitor` data sources, rather than an independent capture layer.
+2. The dashboard can show **multiple live RLCR loops simultaneously**, with
+   per-loop status and streamed log output.
+3. The entry point moves out of Claude (no more `/humanize:viz` slash command)
+   and into the `humanize monitor` CLI command, as a new web-online viewing
+   subcommand.
+4. The new capability targets **online / remote viewing in a browser**, not a
+   local-only viewer that requires the user to be on the same machine running
+   Claude.
+5. Useful features from the existing viz-dashboard branch — notably **cross-
+   conversation querying** (browsing past sessions / loops across different
+   Claude conversations) — are preserved.
+
+## Non-goals
+
+- Reimplementing the monitor capture pipeline (`humanize monitor rlcr/skill/
+  codex/gemini`). The dashboard consumes it; it does not replace it.
+- Continuing to ship `/humanize:viz` as a Claude slash command.
+- Adding chart panels or features explicitly removed in commit 1b575fe
+  ("multi-project switcher + restart + remove chart panels").
+
+## Required behaviors
+
+1. **CLI entry point unification**
+   - Remove `commands/viz.md` and any `/humanize:viz` Claude command surface.
+   - Add a new `humanize monitor` subcommand (name to be agreed during
+     planning, e.g. `humanize monitor web` or `humanize monitor dashboard`)
+     that starts the web dashboard server.
+   - The other `humanize monitor rlcr|skill|codex|gemini` subcommands must
+     keep working unchanged (terminal-attached live tail).
+
+2. **Live multi-loop view**
+   - The web dashboard MUST be able to display 2+ concurrently running RLCR
+     loops at the same time, each with:
+     - current status (running, paused, converged, stopped, …)
+     - current round / phase
+     - live streamed log output, updated in near real time
+
+3. **Reuse existing monitor data**
+   - The dashboard MUST source its data from the same files / events that
+     `humanize monitor rlcr/skill/codex/gemini` already read. It MUST NOT add
+     a parallel capture mechanism (no new hooks just for the dashboard).
+
+4. **Online / remote-viewable**
+   - The dashboard MUST be reachable from a browser over the network, not
+     only via `localhost` on the machine running Claude. Concrete binding /
+     auth design to be agreed during planning.
+
+5. **Cross-conversation history**
+   - Cross-conversation querying (browsing past loops from different Claude
+     conversations / sessions) from the existing viz-dashboard branch MUST be
+     preserved.
+
+## Branch hygiene
+
+Before implementation begins, the branch `feat/viz-dashboard` MUST be rebased
+onto the latest `upstream/dev` (humania-org/humanize). Several relevant changes
+have landed on `upstream/dev` after the branch diverged, including:
+
+- `Add ask-gemini skill and tool-filtered monitor subcommands` (introduces the
+  `humanize monitor skill|codex|gemini` subcommands the dashboard must reuse)
+- `Remove PR loop feature entirely` (the viz-dashboard branch still references
+  PR-loop concepts via `commands/cancel-pr-loop.md`, `commands/start-pr-loop.md`,
+  `hooks/pr-loop-stop-hook.sh`)
+- Multiple monitor / hook fixes
+
+The rebase is therefore both a precondition for correctness (the dashboard
+consumes the new monitor subcommands) and a cleanup step (PR-loop references
+must be dropped).
+
+## Out of scope (for this plan)
+
+- Changes to RLCR semantics, hooks, or skill behavior.
+- Authentication providers, identity systems, or multi-user account models —
+  basic remote-access protection is in scope, but full IAM is not.