MCP server: One MCP server.

Ragmetric

@mukundakattaMCP server · External

RAG retrieval IR metrics: recall@k, hit@k, MRR, NDCG@k, evaluate_batch.

AI evaluation & benchmarksv0.1.1Apache-2.0JavaScript
Freeno charge per call

Runs on your machine from its package; the steps are below.

What it does

  1. Reads
  2. Writes
  3. High impact

At most: only reads or looks up information.

What it costs

Free

Odel charges nothing per call. The publisher may charge for its own service.

Runs locally over stdio

This server isn't hosted — your MCP client launches it from a package registry. Use one of the commands below, or drop the config into your client (e.g. Claude Desktop).

npm@mukundakatta/ragmetric-mcpv0.1.1

Run

npx -y @mukundakatta/ragmetric-mcp

MCP client config

{
  "mcpServers": {
    "ragmetric": {
      "command": "npx",
      "args": [
        "-y",
        "@mukundakatta/ragmetric-mcp"
      ]
    }
  }
}
README3 min read

mcp-stack

Fourteen small MCP servers for everyday LLM / RAG / agent work. Each ships independently on npm, listed in the official MCP Registry, and is callable from Claude Desktop, Cursor, Cline, Windsurf, Zed, and any other MCP-compatible client.

RAG / agent helpers

PackageToolsWhen to install
@mukundakatta/promptbudget-mcptruncate_to_token_budgetLLM needs to fit text into a context window mid-task
@mukundakatta/citecite-mcpinject_citations, parse_citations, strip_citationsRound-trip [1] [2] markers in RAG outputs
@mukundakatta/ragmetric-mcprecall_at_k, hit_at_k, mrr, ndcg_at_k, evaluate_batchLLM helps score retrieval quality
@mukundakatta/ragdrift-mcpinterpret_drift_score, recommend_thresholds, explain_drift_dimensionsLLM diagnoses RAG drift alerts

Reliable transforms LLMs reach for tools instead of imagining

PackageToolsWhen to install
@mukundakatta/csv-tools-mcpparse_csv, to_csv, pluck_columnsLLM needs RFC 4180-correct CSV (quoted fields, embedded commas, BOMs)
@mukundakatta/regex-test-mcptest_regex, find_all, replaceLLM needs exact regex semantics with real match offsets and named groups
@mukundakatta/jmespath-mcpjson_queryLLM needs deep JSON traversal (filters, projections, pipes) without hallucinating
@mukundakatta/diff-mcpunified_diff, apply_patch, parse_patchCode-review or code-edit agent needs character-precise patches
@mukundakatta/sqlfmt-mcpformat_sql, list_dialectsDeterministic SQL formatting across 19 dialects
@mukundakatta/shellquote-mcpquote_bash, quote_bash_argv, quote_cmd, quote_powershellSafe shell argument escaping; LLMs constantly mishandle quotes/$VAR
@mukundakatta/json5-mcpparse_json5, to_json5, to_strict_jsonLLM emitted JSON-with-comments / trailing-commas; round-trip to strict JSON
@mukundakatta/toml-yaml-json-mcpparse, format, convertTOML / YAML / JSON conversion (LLMs especially mishandle TOML)
@mukundakatta/timezone-mcpconvert_tz, now_in, tz_offsetIANA timezone math with real DST rules; LLMs hallucinate offsets
@mukundakatta/html-to-markdown-mcphtml_to_md, extract_textWeb-scraping agents need clean Markdown / text from HTML

Sibling libraries

The first four wrap or re-implement logic from the user's Rust + Python work:

The MCP servers are independent: they implement the logic they need in plain JavaScript so installing them doesn't pull a Rust toolchain.

Install (any of them)

Add to your MCP client config. Example for Claude Desktop:

json
{
  "mcpServers": {
    "promptbudget": { "command": "npx", "args": ["-y", "@mukundakatta/promptbudget-mcp"] },
    "citecite":     { "command": "npx", "args": ["-y", "@mukundakatta/citecite-mcp"] },
    "ragmetric":    { "command": "npx", "args": ["-y", "@mukundakatta/ragmetric-mcp"] },
    "ragdrift":     { "command": "npx", "args": ["-y", "@mukundakatta/ragdrift-mcp"] },
    "csv-tools":    { "command": "npx", "args": ["-y", "@mukundakatta/csv-tools-mcp"] },
    "regex-test":   { "command": "npx", "args": ["-y", "@mukundakatta/regex-test-mcp"] },
    "jmespath":     { "command": "npx", "args": ["-y", "@mukundakatta/jmespath-mcp"] },
    "diff":           { "command": "npx", "args": ["-y", "@mukundakatta/diff-mcp"] },
    "sqlfmt":         { "command": "npx", "args": ["-y", "@mukundakatta/sqlfmt-mcp"] },
    "shellquote":     { "command": "npx", "args": ["-y", "@mukundakatta/shellquote-mcp"] },
    "json5":          { "command": "npx", "args": ["-y", "@mukundakatta/json5-mcp"] },
    "toml-yaml-json": { "command": "npx", "args": ["-y", "@mukundakatta/toml-yaml-json-mcp"] },
    "timezone":       { "command": "npx", "args": ["-y", "@mukundakatta/timezone-mcp"] },
    "html-to-md":     { "command": "npx", "args": ["-y", "@mukundakatta/html-to-markdown-mcp"] }
  }
}

Develop

bash
git clone https://github.com/MukundaKatta/mcp-stack
cd mcp-stack
npm install
npm test --workspaces

License

MIT OR Apache-2.0.

Similar modules

Browse AI evaluation & benchmarks