Hero is now an honest web operator surface for the current stack, not a fake swarm demo.

It serves the operator-critical cards only:

• Hermes
• Factory Droids
• Telegram
• GBrain
• WORKFLOWS
• Alerts

Primary entrypoints:

• UI: http://127.0.0.1:8080/
• Status API: http://127.0.0.1:8080/api/status
• Liveness: http://127.0.0.1:8080/api/healthz
• Readiness: http://127.0.0.1:8080/api/readiness

The rebooted dashboard probes real local evidence instead of replaying old cache theater.

• Hermes: checks local Hermes process activity, gateway/webapi ports, and known profile configs
• Telegram: checks Telegram-facing Hermes profiles plus canon/topic metadata
• GBrain: checks Hermes MCP config, sanitized endpoint reachability, and local brain repo state
• WORKFLOWS: checks the WORKFLOWS repo, key lane directories, file counts, and recent commit state
• Alerts: aggregates any degraded, stale, or down subsystem into one operator-facing surface

If a subsystem is old, partial, or missing, Hero should say so plainly.

The reboot path does not present these as current operator truth:

• fake NATS swarm topology
• synthetic agent registries
• Graphiti-first/Neo4j-first control-plane mythology
• analytics dashboards pretending to be runtime health
• random fallback data for tokens, GitHub, or agents

Legacy files remain in the repo, but they are not the canonical reboot path.

By default, the reboot expects local Rudy-style paths:

• ~/.hermes/config.yaml
• ~/brain
• ~/wiki/queries/grok420system.md
• ~/ORGANIZED/ACTIVE_PROJECTS/ARSENAL/WORKFLOWS

You can override those with env vars:

export HERO_HERMES_CONFIG=~/.hermes/config.yaml
export HERO_BRAIN_ROOT=~/brain
export HERO_TELEGRAM_CANON=~/wiki/queries/grok420system.md
export HERO_WORKFLOWS_ROOT=~/ORGANIZED/ACTIVE_PROJECTS/ARSENAL/WORKFLOWS
export HERO_WEBAPI_PORT=8642
export HERO_WORKSPACE_PORT=3011
export HERO_DROID_BRIDGE_PORT=8645
export HERO_FACTORY_SETTINGS=~/.factory/settings.json

Requirements:

• Python 3
• fastapi
• uvicorn
• jinja2
• PyYAML
• optional but useful local assets:
  • ~/.hermes/config.yaml
  • ~/.hermes/profiles/
  • ~/.factory/settings.json
  • ~/.local/bin/droid
  • ~/brain
  • ~/ORGANIZED/ACTIVE_PROJECTS/ARSENAL/WORKFLOWS
  • ~/wiki/queries/grok420system.md

Start the dashboard in the background:

./launch_web_dashboard.sh start

Other useful commands:

./launch_web_dashboard.sh status
./launch_web_dashboard.sh logs
./launch_web_dashboard.sh stop

For an attached interactive run, use:

./launch_web_dashboard.sh foreground

You can also run the app directly:

python3 web_dashboard.py

Current contract shape:

{
  "overview": {
    "timestamp": "2026-04-22T00:00:00Z",
    "environment": "local-dev",
    "mode": "production",
    "version": "hero-reboot-v1"
  },
  "cards": {
    "hermes": {"status": "healthy", "source": "...", "timestamp": "...", "freshness_seconds": 0, "last_error": null, "details": {}},
    "droids": {"status": "healthy", "source": "...", "timestamp": "...", "freshness_seconds": 0, "last_error": null, "details": {}},
    "telegram": {"status": "healthy", "source": "...", "timestamp": "...", "freshness_seconds": 0, "last_error": null, "details": {}},
    "gbrain": {"status": "healthy", "source": "...", "timestamp": "...", "freshness_seconds": 0, "last_error": null, "details": {}},
    "workflows": {"status": "healthy", "source": "...", "timestamp": "...", "freshness_seconds": 0, "last_error": null, "details": {}},
    "alerts": {"status": "healthy", "source": "hero-aggregator", "timestamp": "...", "freshness_seconds": 0, "last_error": null, "details": {"items": []}}
  },
  "card_order": ["hermes", "droids", "telegram", "gbrain", "workflows", "alerts"]
}

Status semantics:

• healthy: probe succeeded and freshness is acceptable
• degraded: partial evidence or partial failure
• stale: evidence exists but is too old
• down: probe failed and no usable current evidence exists

• docs/plans/2026-04-22-hero-reboot-implementation.md
• docs/dashboard-runtime-contract.md
• docs/dashboard-producers-audit.md
• docs/hero-reboot-thin-slice.md

python3 -m py_compile web_dashboard.py
pytest -q tests/test_web_dashboard_reboot.py tests/test_launcher_contract.py
bash -n launch_web_dashboard.sh

This repo still contains a lot of legacy NATS/analytics/terminal-dashboard material. The rebooted dashboard path is only:

• web_dashboard.py
• web_templates/dashboard.html
• launch_web_dashboard.sh
• tests/test_web_dashboard_reboot.py
• the reboot docs in docs/

High-noise legacy docs that are now explicitly marked as historical context:

• ANALYTICS_GUIDE.md
• NATS_GUIDE.md
• README_UPDATED.md
• cli_agent_briefing.md
• CLAUDE.md

If documentation elsewhere disagrees with those files, trust the reboot path first.