nntrivi2001
c7dbaf2d9f
Two leftovers from the rebrand, done together because they touch the same docs and manifests. **Data directory.** `agent-dashboard` could not simply be renamed in the source: it names a real directory holding the live SQLite database. Startup now moves it once, via `migrateLegacyDataDir()` in `server/lib/claude-home.js` — a same-filesystem rename, so it is instant even at multiple gigabytes and cannot leave two divergent copies. `getDataDir()` stays PURE and resolves to the legacy path until the move happens; the first version of this put the rename inside the getter, which meant `npm run test:server` relocated the developer's real database as a side effect. The call sits behind `require.main === module` and ahead of the router requires, because those pull in db.js, which opens SQLite at module load. A failed rename is silent and the old directory keeps being used: an unfashionable directory name beats a blank database beside an unreachable one. Same treatment for the port-discovery file (`.agent-dashboard.json` → `.agenthub.json`, old one dropped on write). Not renamed on purpose: the Sagi localStorage keys (renaming silently resets everyone's prefs) and the PWA manifest `id` (changing it orphans an installed app). **MCP server.** 4 tool calls in 46 days across 577,132 events, against 4,556 lines of TypeScript, a 2 MB committed build artifact, and a pre-commit guard that had already blocked commits twice. Deleted: `mcp/`, the maintainer skills and `mcp-reviewer` agent, the build check, `plugins/agenthub/.mcp.json`, `docs/MCP.md`, 12 npm scripts, 5 Makefile targets, and every doc section describing it. Kept, because they are unrelated things sharing the name: the Config Explorer's MCP tab and its API, per-lane Playwright MCP, the agenthub-config plugin's MCP audit, and the historical `mcp__plugin_agenthub_*` events that analytics must keep attributing. Verified: server 1432/1432, client 459/459, tsc clean, prettier clean. Both directions of the migration proven — a sandbox-HOME test run leaves the legacy directory untouched, and a real boot moves it with its contents intact.
117 lines
6.3 KiB
Bash
117 lines
6.3 KiB
Bash
# Claude Code home directory (default: ~/.claude)
|
|
# Override this if your Claude Code data is in a different location, e.g.:
|
|
# CLAUDE_HOME=~/.codefuse/engine/cc
|
|
# CLAUDE_HOME=~/.claude
|
|
|
|
# ── Server / network ────────────────────────────────────────────────────────
|
|
# Port to listen on (default: 4820)
|
|
# DASHBOARD_PORT=4820
|
|
|
|
# Directory the built client is served from in production (default: client/dist).
|
|
# The `agenthub` plugin install points this at ~/.claude/agenthub/runtime/client-dist
|
|
# because the plugin cache directory is read-only and replaced on every update.
|
|
# DASHBOARD_CLIENT_DIST=
|
|
|
|
# Interface to bind. SECURITY: defaults to 127.0.0.1 (loopback) so the dashboard
|
|
# is NOT reachable from the network out of the box (GHSA-gr74-4xfh-6jw9). The
|
|
# server reads transcripts, exports all data, and can spawn `claude`, so binding
|
|
# it to the network without auth exposes all of that. Only widen the bind if you
|
|
# understand the risk — and set DASHBOARD_TOKEN when you do.
|
|
# DASHBOARD_HOST=0.0.0.0 # bind all interfaces (LAN-reachable)
|
|
|
|
# Optional auth token. When set, every /api/* request and the WebSocket must
|
|
# present it (Authorization: Bearer <token>, x-dashboard-token header, or
|
|
# ?token=). Strongly recommended whenever DASHBOARD_HOST is non-loopback. Unset
|
|
# by default — the loopback bind is the trust boundary.
|
|
# DASHBOARD_TOKEN=change-me-to-a-long-random-string
|
|
|
|
# Extra Host-header names allowed besides loopback (comma-separated). Needed when
|
|
# you bind to a LAN and reach the dashboard by hostname/IP — entries here pass
|
|
# the anti-DNS-rebinding Host allowlist. Also required when Prometheus-in-Docker
|
|
# scrapes a host-native dashboard:
|
|
# DASHBOARD_ALLOWED_HOSTS=host.docker.internal
|
|
# DASHBOARD_ALLOWED_HOSTS=dashboard.internal,192.168.1.50
|
|
|
|
# ── Background sweeps / sync ─────────────────────────────────────────────────
|
|
# Minutes of inactivity before an active session is marked "abandoned" by the
|
|
# periodic maintenance sweep (default: 180 = 3h).
|
|
# DASHBOARD_STALE_MINUTES=180
|
|
|
|
# Idle-working timeout (seconds) for recovering a turn cancelled with Esc BEFORE
|
|
# any output (which leaves no transcript marker). When the main agent has been
|
|
# "working" with no tool in flight and neither a hook nor the transcript has
|
|
# advanced for this long, the watchdog moves the session to Waiting (default: 120).
|
|
# DASHBOARD_WORKING_IDLE_SECONDS=120
|
|
|
|
# Dead-session liveness reap. The watchdog lists running `claude` CLI processes
|
|
# (ps + lsof on macOS, /proc on Linux) and completes any active session whose
|
|
# cwd has no live claude process — recovering sessions whose SessionEnd hook
|
|
# was lost because the dashboard was down when the user quit (e.g. Ctrl+C).
|
|
# Set to 0 to disable (do this when hooks arrive from another machine); it is
|
|
# auto-disabled on Windows and inside containers (default: enabled).
|
|
# DASHBOARD_LIVENESS_PROBE=1
|
|
|
|
# Idle gate (seconds) for WATCHDOG-TICK liveness reaps: a session is only
|
|
# completed when its transcript hasn't been written for at least this long
|
|
# (the last hook write is the fallback clock when no transcript exists on
|
|
# disk), so mid-turn or just-resumed sessions never flicker out on a transient
|
|
# probe miss. The startup passes skip this gate entirely — at boot the probe
|
|
# alone decides, so sessions quit moments before launch clear immediately
|
|
# (default: 60).
|
|
# DASHBOARD_LIVENESS_IDLE_SECONDS=60
|
|
|
|
# Poll interval (ms) for the continuous ~/.claude/projects sync that surfaces
|
|
# projects added after startup whose sessions never flow through hooks. The
|
|
# fs.watch watcher fires near-instantly regardless; this poll is the safety net.
|
|
# Set to 0 to disable the poll while leaving the watcher running (default: 30000).
|
|
# DASHBOARD_SESSION_SYNC_MS=30000
|
|
|
|
# ── Remote Data Sources (SSH multi-machine collection) ───────────────────────
|
|
# The dashboard can pull Claude Code history from other machines over SSH:
|
|
# it rsyncs each enabled remote's ~/.claude/projects into a sandboxed staging
|
|
# dir, feeds it through the same importer used for local history, and tags the
|
|
# imported sessions with the source. Authentication defers entirely to the host's
|
|
# own SSH stack (~/.ssh/config, ssh-agent, keys, known_hosts) — no passwords or
|
|
# secrets are stored here.
|
|
|
|
# Poll interval (ms) for the background poller that syncs each enabled remote
|
|
# source. Set to 0 to disable the poller (manual / on-demand syncs still work)
|
|
# (default: 15000 = 15s).
|
|
# DASHBOARD_REMOTE_SYNC_MS=15000
|
|
|
|
# Per-source timeout (ms) for a single remote sync (rsync pull + import) before
|
|
# it is aborted (default: 600000 = 10min).
|
|
# DASHBOARD_REMOTE_SYNC_TIMEOUT_MS=600000
|
|
|
|
# Timeout (ms) for the Test SSH probe that verifies a remote source is
|
|
# reachable (default: 15000 = 15s).
|
|
# DASHBOARD_REMOTE_TEST_TIMEOUT_MS=15000
|
|
|
|
# Freshness window (ms) for a remote source session's live status. On each sync,
|
|
# a remote session whose mirrored transcript changed within this window is kept
|
|
# active; once the mirror stops advancing for longer, it is reconciled to
|
|
# completed. Remote sessions get no live hooks, so this replaces the local
|
|
# liveness/stale sweeps (which skip them). (default: 600000 = 10min).
|
|
# DASHBOARD_REMOTE_ACTIVE_WINDOW_MS=600000
|
|
|
|
# ── Lanes (durable parallel-work units and managed git worktrees) ────────────
|
|
# Directory dashboard-managed worktrees are provisioned under. It is also the
|
|
# boundary safety check 2 resolves every destructive lane path against: nothing
|
|
# outside it can be reset or removed. Keep it readable only by the service user
|
|
# (default: ~/.claude/agenthub-lanes).
|
|
# LANES_ROOT=~/.claude/agenthub-lanes
|
|
|
|
# Default base branch a new managed worktree resolves against when the request
|
|
# omits `base`. An explicit --base / request body `base` always wins
|
|
# (default: main).
|
|
# LANE_BASE_BRANCH=main
|
|
|
|
# Prefix for a new managed worktree's feature branch name, as <prefix><slug>
|
|
# (default: feat/).
|
|
# LANE_BRANCH_PREFIX=feat/
|
|
|
|
# Seconds of silence after which a lane whose stage matches /watch|poll/i flips
|
|
# from `active` to `dead` liveness. A silent *idle* lane is at rest, not dead
|
|
# (default: 300 = 5min).
|
|
# LANE_DEAD_SEC=300
|