From LoreConvo
First-time setup wizard for LoreConvo that verifies MCP server connectivity, database access, hooks configuration, and runs a test save/load cycle.
How this skill is triggered — by the user, by Claude, or both
Slash command
/loreconvo:lore-onboardThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Welcome the user and walk them through first-time LoreConvo setup verification.
Welcome the user and walk them through first-time LoreConvo setup verification. Run each step below in order. Report results as you go. If any step fails, explain what went wrong and how to fix it before continuing.
Call the get_recent_sessions MCP tool with limit=1.
If FAIL: Tell the user:
LoreConvo MCP server is not connected. Make sure the plugin is installed and the MCP server entry is in your Claude settings. Check that
~/.loreconvo/exists and the server can be reached via stdio transport.
Call list_projects MCP tool.
If FAIL: Tell the user:
The LoreConvo database could not be accessed. Check that
~/.loreconvo/sessions.dbexists and is not locked by another process. You can reset it by deleting the file and restarting -- LoreConvo will recreate it automatically.
Call save_session with:
title: "LoreConvo Onboarding Test"
surface: "code"
summary: "Automated onboarding verification. This session confirms that LoreConvo can save and retrieve sessions correctly. Safe to delete."
decisions: ["LoreConvo onboarding test completed successfully"]
tags: ["onboarding", "test"]
Capture the returned session_id.
Call get_session with the session_id from step 3a.
Call search_sessions with query "onboarding test".
Check whether the auto-save and auto-load hooks are configured by looking for the hook files. Use the Bash tool to check:
ls -la "${CLAUDE_PLUGIN_ROOT:-$HOME/.claude/plugins/loreconvo}/hooks/scripts/"
Or check the plugin.json for SessionStart and SessionEnd hooks.
on_session_start.sh and on_session_end.sh exist.If FAIL: Tell the user:
Hooks are not configured. The auto-save (SessionEnd) and auto-load (SessionStart) hooks make LoreConvo work automatically. Reinstall the plugin or check that hooks/scripts/ contains on_session_start.sh and on_session_end.sh.
Check if the user's project CLAUDE.md (in the current working directory) mentions LoreConvo. Read the CLAUDE.md file if it exists.
If SUGGEST, offer to append this snippet:
## LoreConvo Integration
This project uses LoreConvo for persistent session memory.
- At session start: context from recent sessions is auto-loaded via hooks
- At session end: the session is auto-saved via hooks
- Use `save_session` to manually vault important context mid-session
- Use `search_sessions` or `get_context_for` to recall prior work
- Use `vault_suggest` for proactive recommendations on what to revisit
Ask the user before modifying their CLAUDE.md.
After all steps, present a summary table:
| Step | Check | Result |
|---|---|---|
| 1 | MCP Server Connection | PASS/FAIL |
| 2 | Database Access | PASS/FAIL |
| 3a | Save Session | PASS/FAIL |
| 3b | Load Session | PASS/FAIL |
| 3c | Search (FTS5) | PASS/FAIL |
| 4 | Hooks Configured | PASS/FAIL |
| 5 | CLAUDE.md Integration | PASS/SKIP/SUGGEST |
If all critical steps (1-4) pass, congratulate the user:
LoreConvo is fully operational! Your sessions will be automatically saved and loaded. Use
/vault saveor ask me to "vault this session" anytime to manually capture important context.
If any critical step fails, summarize what needs fixing and offer to help.
Do NOT delete the test session -- it serves as the user's first session record and confirms the installation date. The "onboarding" tag makes it easy to find later if the user wants to clean it up.
npx claudepluginhub labyrinth-analytics/claude-plugins --plugin loreconvoGuides completion of development work by verifying tests, detecting environment, and presenting structured options for merge, PR, or cleanup.
Guides creation and editing of skills using test-driven development with pressure scenarios and subagents to verify agent compliance.
Dispatches multiple subagents concurrently for independent tasks without shared state. Use when facing 2+ unrelated failures or subsystems that can be investigated in parallel.