vector-hybrid-search

vector-hybrid-search is a skill for Claude Code, Codex from Knuckles-Team/vector-mcp. It costs 113 tokens per session (931 once invoked), scanned A, original, MIT.

A search tool for an existing vector-store collection, which is a database organized to find similar pieces of text. It supports meaning-based search, exact-word search with BM25, or a hybrid ranking that combines both.

In plain words
What is it for?
Use it to fetch top matching text chunks for retrieval-augmented generation, answer questions from an indexed corpus, find exact identifiers or code terms, or combine semantic and keyword results.
Why use it?
It helps retrieve relevant passages from an indexed collection when ordinary keyword matching misses related wording, or when meaning-based search misses exact codes and rare terms. The collection must already exist and contain data.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Use it to fetch top matching text chunks for retrieval-augmented generation, answer questions from an indexed corpus, find exact identifiers or code terms, or combine semantic and keyword results.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/knuckles-team/vector-mcp/vector-hybrid-search
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add Knuckles-Team/vector-mcp --skill vector-hybrid-search
Clone the repo
git clone --depth 1 https://github.com/Knuckles-Team/vector-mcp

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for vector-hybrid-search

README.md
[![agentmods](https://agentmods.dev/badge/skills/knuckles-team/vector-mcp/vector-hybrid-search/github.svg)](https://agentmods.dev/skills/knuckles-team/vector-mcp/vector-hybrid-search)
Your own site
<a href="https://agentmods.dev/skills/knuckles-team/vector-mcp/vector-hybrid-search"><img src="https://agentmods.dev/badge/skills/knuckles-team/vector-mcp/vector-hybrid-search/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for vector-hybrid-search

Your own site · 80×15
<a href="https://agentmods.dev/skills/knuckles-team/vector-mcp/vector-hybrid-search"><img src="https://agentmods.dev/badge/skills/knuckles-team/vector-mcp/vector-hybrid-search.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 113 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 931 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00113 $0.00931
Opus 5 $0.00056 $0.00465
Sonnet 5 $0.00023 $0.00186
Haiku 4.5 $0.00011 $0.00093

Measured 8d ago against content hash f650605483de, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

vector-hybrid-search scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

vector_mcp/skills/vector-hybrid-search/SKILL.md · 84 lines

How it starts

The opening of the file, as written. The whole thing — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Retrieval over vector-store collections through the vector-mcp MCP server. One condensed tool exposes three retrieval strategies: dense semantic search, sparse lexical BM25 search, and a hybrid that fuses both rankings.

When to use

  • Answer a natural-language question from an indexed corpus (grab RAG context).
  • Fetch the top-k most relevant chunks for a query.
  • Do keyword/term-exact retrieval where BM25 beats embeddings (codes, IDs, rare tokens).
  • Balance recall (semantic) against precision on exact terms (lexical) via hybrid fusion.

When NOT to use

  • Creating / populating / deleting collections → vector-collection-management.
  • Selecting or configuring the backend engine → vector-backend-operations.
  • General web search or KG queries — this only searches the named collection.

Prerequisites & environment

Connect via the mcp-client skill against the vector-mcp MCP server. The collection must already exist and be populated (see vector-collection-management). Connection params (db_type, db_path, host/port, db_name, username, password, collection_name) are passed inline, matching the collection you want to query.

Tools & actions

Condensed tool Actions
vector_search semantic_search, lexical_search, search (hybrid)

Key parameters

  • question — the query text (required for every action).
  • number_results — top-k to return.
  • semantic_weight / bm25_weight — leg weights for the hybrid search action.
  • rrf_k — the Reciprocal Rank Fusion constant that merges the two rankings.

Recipes

Semantic (vector) search:

{"action":"semantic_search","db_type":"chromadb","db_path":"./chroma","collection_name":"handbook","question":"how do I connect to the VPN?","number_results":5}

Lexical BM25 search (term-exact):

{"action":"lexical_search","db_type":"chromadb","db_path":"./chroma","collection_name":"handbook","question":"gw-hq","number_results":5}

Hybrid search, semantic-leaning fusion:

{"action":"search","db_type":"chromadb","db_path":"./chroma","collection_name":"handbook","question":"vpn gateway hostname","number_results":8,"semantic_weight":0.7,"bm25_weight":0.3,"rrf_k":60}

Read the full file on GitHub · 84 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 8d ago First seen · 84 lines · 113 tokens per session scan A f650605483de

Subscribe to this mod's changes

vector-hybrid-search is a skill published in the GitHub repository Knuckles-Team/vector-mcp (15 stars, last pushed 11d ago), licensed MIT. It adds 113 tokens to every session and 931 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.