"A local vector-based search engine for personal documents" matching MCP tools:

• nexus_similarity_search_api_rank_items_by_nmi_cosine_fusion

  similarity-search-api-sdk

  Ranks a corpus of items against a query vector using a calibrated fusion score (alpha * cosine + (1-alpha) * NMI_normalizado), where alpha is auto-derived from the corpus's marginal entropy unless overridden. Results are identified by their 0-indexed position in corpus_vectors (this tool does not accept explicit item IDs). Use this when you need semantically-calibrated similarity over a stateless corpus of up to 500k items without a vector database. Do NOT use for purely geometric nearest-neighbor search where NMI overhead is unnecessary, nor for corpora larger than 500k items per call. Requires a valid api_key (same as X-API-Key) and an x402 payment.

  Connector

• collection.search

  DocImprint

  Semantic (vector) search across documents in a collection. Returns ranked text chunks with relevance scores. Free — no credits consumed. Use when you need raw matching chunks from a collection. For a synthesized cited answer from the same context, use collection.ask instead. PREREQUISITE: Collection must be populated via collection.add_document and async indexing must complete (poll job.status) before results appear. Returns: { results: [{ bundle_id, chunk_id, text, score: number (0–1), title? }] } Example prompts: - "Search my Q4 Contracts collection for mentions of liability cap." - "Find the clause about data retention in my due diligence docs." - "Search for revenue numbers across my quarterly reports."

  Connector

• collection.ask

  DocImprint

  Answer a question using RAG over a document collection. Retrieves relevant chunks then synthesizes a cited answer with source attribution. Use when you need a direct answer grounded in your collection documents. For raw matching chunks (without synthesis), use collection.search instead. For single-document Q&A, use url.qa instead. PREREQUISITE: Collection must be populated via collection.add_document and indexed before results appear. Returns: { answer: string, sources: [{ bundle_id, chunk_id }], retrieval: [{ bundle_id, chunk_id, text, score }] } Example prompts: - "What are the key terms of the service agreement in my collection?" - "Based on my due diligence docs, what are the main risks?" - "Answer this question using all documents in the Q4 Contracts collection."

  Connector

• semantic_search

  canton-ccpedia

  PRIMARY tool for open-ended questions: how / why / what-is, troubleshooting a symptom ("why is my balance zero", "how do I fix X"), and locating config or setup steps. Conceptual/meaning-based search over the full Canton corpus (CIPs, docs, forum, mailing lists, proposals, blog, releases, ecosystem, foundation KB, YouTube) using vector+FTS hybrid retrieval with reranking. Canton-specific. Use this FIRST for anything a specific tool does not clearly own; the narrow curated tools (get_faq, find_known_issues, diagnose_error) cover only small hand-picked sets or need a literal error string, so prefer semantic_search for real how/why/config questions. Then call get_doc with a returned id to read the full source page.

  Connector

• build_knowledge_index

  agent4.io

  Build (or rebuild) the structured index for a knowledge base — the second leg beside vector search. Vector search answers "what does this passage say". It **cannot count, filter numerically or aggregate**, so "how many documents", "which ones are between 1000 and 2000 words", "how many per category" are not answered badly — they are structurally unanswerable. This builds a small per-KB table from whatever structured header the documents share, which the agent can then query with SQL via `query_knowledge_table`. Only worth it when the documents share a machine-readable header (a metadata table, YAML front matter, `Field: value` lines). **Prose gets declined, and that is the right answer** — a table of unique values makes statistics meaningless. `roles` names the fields that must be extracted **exactly** and never paraphrased. Use it when the answer has to quote something the model must not invent: - `identity` — what to call the item (book title, drug name, product name) - `link` — where to send the user - `image` — what to show the user - `code` — the unique identifier Which link is "the" link is a business fact the data does not state — only the customer knows. A declared role that cannot be found comes back in `roles.unresolved` **with candidate field names**: ask the user which one it is, do not guess. **Read `dropped` in the report and tell the user about it.** A column that was thrown out (coverage too low, two columns holding identical values) is invisible in later query results — the model simply works around it — so this report is the only place it is ever mentioned.

  Connector

• create_campaign

  xmagnet

  Create a DRAFT email campaign via a programmatic wizard. Call this tool and it will guide through the steps — no manual orchestration needed. WIZARD STEPS (handled automatically by the tool): 1. Call with contacts + total_contacts → tool returns engine picker (NextGen vs MyConvo) 2. Add campaign_type from user's click → tool returns campaign category chips (promotional, newsletter, event…) 3. Add campaign_category from user's click → tool returns engine-specific template gallery MyConvo: shows plain_email_templates (personal plain-text). NextGen: shows campaign_templates (HTML). 4. Add template_id from user's pick → tool creates the draft campaign. RULES: Reuse contacts from prior search — never re-search. Pass total_contacts from search result's total_in_crm so the user always sees the full count. Saves as DRAFT only — no emails sent.

  Connector

• get_location_options

  Loppee

  Return only the calling customer's safe Home metadata plus opaque saved_location_ref, current-location handoff availability, and provided city/ZIP context. Never returns ZIP, address, latitude, or longitude. Ask exactly: "Where should I search? For local categories, share your current location or use Home only when it is marked exact. Another city/ZIP is for area context or business-name lookup only." Only current or Home with kind=saved_exact authorizes local category/recommendation discovery; provided city/ZIP and coarse Home are neutral name-disambiguation context. Requires an active scoped consumer personal-agent key.

  Connector

• record_execution

  Routescore

  Persist 1-100 already-built execution_record.v1 documents as immutable, personal-account post-trade evidence. Requires a stable idempotency_key; an exact replay returns the original result and reuse with different content fails closed. Pro tier, owner-scoped, and bounded by the MCP transport byte limit. This tool never accepts calldata, signing material, routes funds, executes a transaction, applies reviewer overrides, or exposes another tenant. Invalid rows are recorded only as redacted rejection reasons and are never echoed.

  Connector

• find_agents_for_task

  AgentMarketplace

  Ranked candidate agents for a task, before any bid exists. Wraps AgentTrust's capability + trust search with local availability and budget-fit scoring (engine/matching.py, PLAN.md §7.1). Use ``evaluate_bids`` instead once actual bids have come in. Args: access_token: AgentAuth bearer token (requires ``market.read``). task_id: UUID of the task to find candidates for. limit: Maximum candidates to return, 1-50. Default 10. Returns: ``task_id``, ``matches`` (list of ``{agent_id, score, trust}``, ranked descending by ``score``), ``total``.

  Connector

• generate_policies

  lexvibe

  Generate the legal documents (privacy policy, terms of service and, if applicable, an AI disclosure) localized and tailored to the target markets (GDPR, UK GDPR, CCPA…). Returns Markdown drafts. Pass check_website's or check_store's suggestedAnswers as `answers` so the documents disclose the right processing. Anonymous remote generation is template-based and capped at 3 locales; AI-tailored, hosted and auto-updated documents require a LexVibe account (https://golexvibe.com).

  Connector

• kirk_infer_legacy

  Kirk — Unsupervised Structural Change Detection

  Score a 50-value feature vector against the legacy /v1/infer route on the sealed engine. Purpose: Backwards-compatible scoring surface for callers that were already targeting the legacy path. Use when: You have an existing client wired to /v1/infer and need continued MCP access without refactoring. Do not use when: You are on a fresh integration — prefer kirk_score_book (single-layer, cascade-shaped path). Also do not use in a tight loop against a large corpus: the MCP round-trip is millisecond-scale, and the LLM tool-call cost accrues per book for agent-driven callers. For bulk work, call kirk_bulk_howto first. Capability class(es): C2 (cross-section entropy scoring), legacy interface. Path fit: Validation via MCP (this tool). Production integrations run in-process under sealed-engine attestation — same binary sha as this endpoint. Contact Kavara for deployment options. Cost: 1 IU per call. For agent-driven callers, per-call LLM tokens accrue on top; the response _cost envelope surfaces both.

  Connector

• get_security_writing_guidelines

  website-search

  Get Lenny Zeltser's expert writing guidelines for security reports and assessments. Provides guidance on tone, structure, clarity, executive summaries, and avoiding common writing mistakes. Includes rating-sheet items (the four lens sheets: structure, look, words, tone) as concrete reference points for grounded feedback. Works for any security document. This server never requests your documents and instructs your AI to keep them local—guidelines flow to your AI for local analysis. Note: For incident response reports specifically, use the ir_* tools which provide deeper section-by-section review criteria.

  Connector

• malware_get_template

  website-search

  Get Lenny Zeltser's malware analysis report template. The report covers Executive Summary, Sample Snapshot, Malware Family Identification, Component Inventory, Runtime Requirements, Sources, Capabilities, Indicators of Compromise, Analysis Details, What We Don't Know, optional Infection Vector, optional Detection Engineering, About this Report, Appendix: Analysis Environment, and optional Appendix: Analysis Scripts. This server never requests your sample, analysis notes, or indicators and instructs your AI to keep them local—guidelines and the report template flow to your AI for local analysis.

  Connector

• search_docs

  pg-aiguide

  Search documentation with hybrid semantic (vector) and keyword (BM25) search. Use semanticWeight to choose keyword-only (0), semantic-only (1), or a blend; mid values fuse rankings with RRF. Supports Tiger Cloud (TimescaleDB), PostgreSQL, and PostGIS.

  Connector

• create_location_handoff

  Loppee

  Create a short-lived current precise-location request with a browser fallback. Once approved, it can authorize local category/recommendation reach and Sponsored exposure. Follow next_action exactly: poll only for poll_location_handoff; for ask_user_to_open_location_url, show the fallback immediately and explain that the user can enable Location for your assistant once under Account > My external AI agent in the Loppee iPhone app so future requests complete automatically. The assistant receives readiness and normal business results, never latitude or longitude. Requires an active database-backed customer personal-agent key.

  Connector

• call_engine

  reefapi-mcp

  Call a ReefAPI engine action — POST /<engine>/v1/<action> with `params`. Returns the uniform { ok, data, meta, error } envelope. Get param names from get_engine_schema first. Needs YOUR ReefAPI key (the local server reads REEFAPI_KEY; the hosted server reads the `Authorization: Bearer ak_live_...` header you configure on the connection). Get a key at https://reefapi.com. Failed calls cost no credits.

  Connector

• get_catalog

  reefapi-mcp

  The FULL ReefAPI catalog — EVERY engine with its one-line title, grouped by category. This is the whole menu (≈ a few thousand tokens); SCAN IT AND PICK THE BEST ENGINE YOURSELF. You are an LLM, so you match the user's intent semantically — across ANY language, typo, or phrasing — far better than a keyword search can. Use this whenever search_engines didn't surface the right engine (or to be sure you didn't miss a better one). After you pick: get_engine_schema(engine) -> get_action_schema -> call_engine.

  Connector

• zoom_list_recordings

  local-mcp

  Lists Zoom meeting recordings saved locally on this Mac (~/Documents/Zoom), newest first: meeting name, date, and which artifacts exist (transcript, captions, saved chat, audio, video). Local recordings only — no Zoom API, no admin approval. Use zoom_read_transcript to read the text of a meeting.

  Connector

• verify_identity

  OpenWarrant — Document Verification Suite

  Run an Australian identity check over a SET of identity documents. A vision model reads each document (which ID it is, which fields it shows — name/photo/address/signature — and its issue date); a deterministic engine then tallies them against a scheme and reports whether identity is established, and exactly what's still missing if not. USE THIS WHEN someone needs to verify a person's identity from their documents — KYC / onboarding / "do these documents satisfy the 100-point check?" Pass ALL the person's documents together (a passport alone is 70 points; the check needs >= 100). `documents` is a list, each item ONE of: {"url": "https://..."} (public link, fetched server-side) or {"bytes_b64": "...", "filename": "passport.pdf"} (inline). Up to 10. `scheme`: "afp_100_point" (points, default) or "austrac_safe_harbour" (category combinations). Returns `{established, points/target or satisfied_path, documents[] (per-document: type, fields shown, whether it counted and why-not), reason, accepts, ...}`. This is identity COVERAGE, not a forgery judgment — run verify_document for authenticity. Documents are never stored.

  Connector

• collection.ask

  api

  Answer a question using RAG over a document collection. Retrieves relevant chunks then synthesizes a cited answer with source attribution. Use when you need a direct answer grounded in your collection documents. For raw matching chunks (without synthesis), use collection.search instead. For single-document Q&A, use url.qa instead. PREREQUISITE: Collection must be populated via collection.add_document and indexed before results appear. Returns: { answer: string, sources: [{ bundle_id, chunk_id }], retrieval: [{ bundle_id, chunk_id, text, score }] } Example prompts: - "What are the key terms of the service agreement in my collection?" - "Based on my due diligence docs, what are the main risks?" - "Answer this question using all documents in the Q4 Contracts collection."

  Connector