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.
2.0 KiB
description
| description |
|---|
| Pull the latest agenthub plugin version and refresh the dashboard runtime |
Pull the newest agenthub release from its marketplace, then reinstall the runtime
dependencies and restart the dashboard server against it. Use this to get a new
dashboard build, or when /agenthub-doctor reports missing dependencies or a dead
server.
First, find which marketplace this install's agenthub came from and pull the
latest version into the plugin cache:
INSTALLED=$(claude plugin list 2>/dev/null | grep -o 'agenthub@[^ ]*' | head -1)
MARKETPLACE="${INSTALLED#agenthub@}"
if [ -z "$MARKETPLACE" ]; then
echo "agenthub is not installed as a plugin (checkout install?) - nothing to update this way."
else
claude plugin marketplace update "$MARKETPLACE"
claude plugin update "agenthub@$MARKETPLACE" -y
fi
That downloads the new version into ~/.claude/plugins/cache/ but does not
take effect in this session — Claude Code only re-resolves ${CLAUDE_PLUGIN_ROOT}
on /reload-plugins (a REPL-only action, not scriptable from a command). Say
so plainly: tell the user to run /reload-plugins now, then either
re-run /agenthub-update (this time the bootstrap step below runs against the new
version) or just start a new session — plugin-bootstrap.js compares the
recorded state against the current plugin version on every SessionStart and
rebuilds automatically if they differ.
Then, whether or not a plugin update was just pulled, refresh the runtime
against whatever ${CLAUDE_PLUGIN_ROOT} currently resolves to. This stops the
running dashboard server before starting the new one. Say so, then run:
node "${CLAUDE_PLUGIN_ROOT}/scripts/plugin-bootstrap.js" --force
It runs in the foreground and can take a few minutes on a cold cache. When it finishes, verify:
node "${CLAUDE_PLUGIN_ROOT}/scripts/plugin-doctor.js"
Report the final state in a few lines. If the bootstrap failed, quote the
decisive line from ~/.claude/agenthub/runtime/bootstrap.log.