ros/n8n-mcp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: optimize MCP tool descriptions for 65-70% token reduction

Browse Source 
- Reduced average description length from 250-450 to 93-129 chars
- Documentation tools now average 129 chars per description
- Management tools average just 93 chars per description
- Moved detailed documentation to tools_documentation() system
- Only 2 tools exceed 200 chars (necessarily verbose)

Also includes search_nodes improvements:
- Fixed primary node ranking (webhook, HTTP Request now appear first)
- Fixed FUZZY mode threshold for better typo tolerance
- Removed unnecessary searchInfo messages
- Fixed HTTP node type case sensitivity issue

This significantly improves AI agent performance by reducing context usage
while preserving all essential information.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
This commit is contained in:
czlonkowski
2025-07-10 11:42:23 +02:00
parent 99e74cf22a
commit d8e84307e4
6 changed files with 640 additions and 187 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
docs/CHANGELOG.md
View File

@@ -5,6 +5,34 @@ All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
## [2.7.11] - 2025-07-10
### Enhanced
- **Token Efficiency**: Significantly reduced MCP tool description lengths for better AI agent performance
- Documentation tools: Average 129 chars (down from ~250-450)
- Management tools: Average 93 chars (down from ~200-400)
- Overall token reduction: ~65-70%
- Moved detailed documentation to `tools_documentation()` system
- Only 2 tools exceed 200 chars (list_nodes: 204, n8n_update_partial_workflow: 284)
- Preserved all essential information while removing redundancy
### Fixed
- **search_nodes Tool**: Major improvements to search functionality for AI agents
- Primary nodes (webhook, httpRequest) now appear first in search results instead of being buried
- Fixed issue where searching "webhook" returned specialized triggers instead of the main Webhook node
- Fixed issue where searching "http call" didn't prioritize HTTP Request node
- Fixed FUZZY mode returning no results for typos like "slak" (lowered threshold from 300 to 200)
- Removed unnecessary searchInfo messages that appeared on every search
- Fixed HTTP node type comparison case sensitivity issue
- Implemented relevance-based ranking with special boosting for primary nodes
### Added
- **FTS5 Full-Text Search**: Added SQLite FTS5 support for faster and more intelligent node searching
- Automatic fallback to LIKE queries if FTS5 is unavailable
- Supports advanced search modes: OR (default), AND (all terms required), FUZZY (typo-tolerant)
- Significantly improves search performance for large databases
- FUZZY mode now uses edit distance (Levenshtein) for better typo tolerance
## [Unreleased]
### Fixed
@@ -516,6 +544,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- Basic n8n and MCP integration
- Core workflow automation features
[2.7.11]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.10...v2.7.11
[2.7.10]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.8...v2.7.10
[2.7.8]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.5...v2.7.8
[2.7.5]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.4...v2.7.5