In-depth architectural comparison of the Conversation Handoff MCP and Scrivener MCP MCP servers. Compare execution transports, security boundaries, tool capabilities, quality scores, and ready-to-paste client installation snippets for Claude, Cursor, Windsurf, and VS Code.
At a Glance & Executive Verdict
Conversation Handoff MCP
Knowledge & Memory · Local stdio
Quality: 60/100 (Good) | Auth: No auth required
Scrivener MCP
Knowledge & Memory · Local stdio
Quality: 61/100 (Good) | Auth: No auth required
Verdict Summary: Choose Conversation Handoff MCP if you need specialized Knowledge & Memory tools running via a local process. Choose Scrivener MCP if your workspace requires Knowledge & Memory integration with local subprocess execution. Both servers can be configured concurrently in your client's mcpServers manifest.
Which MCP Server Should You Choose?
Choose Conversation Handoff MCP when:
You need dedicated capabilities in the Knowledge & Memory domain.
You prefer local stdio subprocess transport architecture.
Your security boundary fits: No auth required (Free / Open Source).
Hand off conversation context between Claude Desktop projects and across MCP clients. Memory-based, no file clutter.
Connect Scrivener 3 writing projects to Claude and other AI assistants. 47 tools for document management, writing analysis, semantic search, character/plot memory, and content enhancement. Progressive skill loading, relationship engine with HMS triplets, and JS fallback for offline semantic search. npm i -g scrivener-mcp
Category & Scope
Tools & Capabilities Breakdown
Conversation Handoff MCP Tools (11)
handoff_save
Save a conversation handoff for later retrieval. Use this to pass conversation context to another AI or project.
## Long conversations (>~100KB)
If the conversation is large, **save the first chunk with this tool**, then call `handoff_append` repeatedly with the remaining chunks. Doing one giant `handoff_save` can fail with an XML parse / tool-input error because Claude Code encodes tool arguments as XML internally and very large strings can break that encoding.
## Format Selection
- **structured** (default): Organize content using the template below. Much faster — reduces output tokens to ~5-20% of the original conversation. Best for most handoffs.
- **verbatim**: Save the complete word-for-word conversation. Use only when exact wording matters (e.g., legal text, precise error messages).
## Structured Template (for format="structured")
```
## Key Decisions
- [Decision]: [Rationale]
## Implementation Details
[What was built/changed, with relevant code snippets]
## Code Changes
[Files modified with brief description]
## Open Issues
- [Issue]: [Status/Context]
## Next Steps
- [ ] Action item
```
Omit sections that don't apply. Add custom sections if needed.
handoff_list
List all saved handoffs with summaries. Returns lightweight metadata without full conversation content. Opens interactive UI if supported.
handoff_load
Load a specific handoff by key. Returns full conversation content.
Ready-to-Paste Client Configurations
Paste either (or both) of these JSON server blocks into your client config file (e.g. claude_desktop_config.json or ~/.cursor/mcp.json).
Conversation Handoff MCP is categorized under Knowledge & Memory and uses a local stdio subprocess. In contrast, Scrivener MCP belongs to Knowledge & Memory using local stdio subprocess. Select Conversation Handoff MCP when you need capabilities focused on knowledge & memory and Scrivener MCP when you require tools for knowledge & memory.
Clear handoffs. If key is provided, clears only that handoff. Otherwise clears all.
handoff_stats
Get storage statistics and current limits. Useful for monitoring usage.
handoff_restart
Restart the shared HTTP server. Useful when the server is in an unhealthy state. All stored handoffs will be lost (data is in-memory).
handoff_merge
Merge multiple handoffs into one. Combines conversations and metadata from related handoffs into a single unified handoff.
handoff_add_comment
Add a comment or annotation to an existing handoff. Comments are included when loading the handoff.
handoff_delete_comment
Delete a comment from a handoff by its comment ID.
handoff_append
Append a conversation chunk to an existing handoff. Use this to upload long conversations in multiple pieces.
## When to use
- The full conversation is over ~100KB.
- handoff_save fails with an XML parse / tool-input error on a long conversation.
Claude Code encodes tool-call arguments as XML internally; very large `conversation` strings can break that encoding (e.g., when the payload contains tag-like substrings). Splitting the upload into smaller chunks avoids that failure mode.
## How to use
1. Call `handoff_save` first with the initial portion (and all metadata: title, summary, etc.) to create the handoff.
2. Call `handoff_append` repeatedly with subsequent chunks until the full conversation is uploaded.
3. Chunks are concatenated as-is — include any newlines or separators inside each chunk.
The cumulative conversation size must stay within the storage limit (default 1 MiB).
handoff_search
Search handoffs by tags, text query, project, AI, status, or date range. Returns matching handoff summaries without full conversation content. Use this to discover relevant handoffs in multi-agent workflows.