context-mode and dynatrace-mcp
These are complements: a context virtualization layer (A) that abstracts tool access through MCP can work alongside an observability platform's MCP server (B) to provide isolated, privacy-protected access to Dynatrace monitoring data within Claude or other AI applications.
About context-mode
mksglu/context-mode
Privacy-first. MCP is the protocol for tool access. We're the virtualization layer for context.
Operates as an MCP server that intercepts tool outputs at the protocol layer, sandboxing large responses (56 KB Playwright snapshots, API logs) into local SQLite databases instead of flooding the context window—achieving 98% token reduction. Implements full-text search (FTS5/BM25) to retrieve only relevant session events during context compaction, preserving conversation continuity without re-injecting raw data. Integrates with Claude Code, Gemini CLI, and VS Code Copilot via hooks that automatically route tool calls, with zero cloud dependencies or telemetry.
About dynatrace-mcp
dynatrace-oss/dynatrace-mcp
MCP server for Dynatrace Observability
Exposes Dynatrace observability data—problems, vulnerabilities, logs, metrics, and spans—through MCP tools for AI assistants, using stdio transport with OAuth browser-based authentication. Includes DQL (Dynatrace Query Language) execution, entity lookups, document management, and Slack/workflow automation, with built-in Grail budget tracking to manage query costs. Integrates directly into VSCode, Cursor, Claude, and other MCP-compatible clients.
Scores updated daily from GitHub, PyPI, and npm data. How scores work