Query Markup Documents - On-device hybrid search for markdown files with BM25, vector search, and LLM reranking
Daftari is installed from its publisher's own source and answers where it runs, so this marketplace is not in the path of a single call. There is no address here to send one to, and a panel that pretended otherwise would be showing you an answer we made up. Install it and call it from your own client — the Installation tab has the entry for each one.
What it does
An open-source, multi-user knowledge vault exposed to AI agents via an MCP server.
Quickstart
# 1 — add it to Claude Code
claude mcp add --scope user daftari-mcp '--' npx -y daftari
# 2 — ask your agent for something these tools do
# vault_read, vault_index, vault_status and 43 more
Daftari costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.
Collected from a public index. Nobody has claimed this account, so nothing here was written by its author — claim it if it is yours.
Where are you running it?
Every route below installs the same thing and ends at the same approval screen. This one runs on your machine: your client starts Daftari as a process under your own user, with your files and your network, so the tool surface below is what it can do to your computer rather than to a server somewhere else. Its scan result is on the Security tab — read the surface before you approve it.
This is a public server: you run it yourself and this marketplace is not in the path. You need Node.js — the current LTS (24) is the safe choice; npx comes with it. Claude Code registers it in one command; --scope user makes it available in every project. The '--' is in quotes so that PowerShell passes it on.
claude mcp add --scope user daftari-mcp '--' npx -y daftariReconnect, or start a new session, and the tools appear in the model’s tool list.
One config entry your client uses to start the process locally. A local server runs with your file system and your network, which is why it is priced without metering.
46 tools, with what each one reads, writes and reaches shown before you agree — the same list on every route above. Read the tool surface.
Tool surface
What the model actually sees. Descriptions are diffed on every release — see version history.
Read from its source code. A server that builds its tool list when it starts can list more tools than this.
Read a single vault document.
No parameter schema on file.
List vault documents with their metadata, including each document's questions_answered / questions_raised.
No parameter schema on file.
Vault health dashboard: the running daftari version, total file count, per-collection counts, count of documents with invalid frontmatter, a staleness distribution (fresh/aging/stale), unresolved tensions, and recent write history.
No parameter schema on file.
Compile an epistemic receipt for the vault documents an answer relies on.
No parameter schema on file.
Compute settled vs contested belief over a depth-bounded topic graph seeded at seed.
No parameter schema on file.
List the end of every supersede chain: per chain, the current entry with its status and date, and the entries it replaced.
No parameter schema on file.
Per-principal track records aggregated from the vault's own ledgers (provenance log, tension log, staged actions), priced by the wager schedule: writing a claim at a confidence level stakes points (low 0 / medium 1 / high 3, provisional); a claim later corrected or retired burns the stake; a claim…
No parameter schema on file.
Hybrid search across the vault: BM25 lexical ranking combined with vector semantic similarity.
No parameter schema on file.
Find documents related to a given vault document.
No parameter schema on file.
Rebuild the search index from the markdown files on disk.
No parameter schema on file.
Surface thematic clusters across the vault using k-means over CHUNK embeddings, with per-document theme distributions.
No parameter schema on file.
For each open question (a questions_raised entry whose exact text is in no document's questions_answered), list the documents touched since it was raised whose title or questions_answered read closest in meaning.
No parameter schema on file.
Create a new vault document or overwrite an existing one.
No parameter schema on file.
Change named frontmatter fields and exact-match body spans in an existing document without resending it.
No parameter schema on file.
Append a markdown section to an existing vault document.
No parameter schema on file.
Promote a draft document to canonical status.
No parameter schema on file.
Mark a document deprecated.
No parameter schema on file.
Change only a document's confidence level (low | medium | high), leaving its status and body untouched.
No parameter schema on file.
Change only a document's write-protection tier (source | compiled | manual), leaving its body untouched.
No parameter schema on file.
Mark a document superseded by a named successor.
No parameter schema on file.
Reverse a supersession: clear superseded_by and return a superseded document to draft, logged as its own provenance action.
No parameter schema on file.
Combine two source documents into a target and supersede both sources to point at it, all in one commit.
No parameter schema on file.
Assert, dispute, or qualify the calling principal's position on an existing claim document.
No parameter schema on file.
Query positions: all positions on one doc ('path'), or all live positions held by a principal across the vault ('principal').
No parameter schema on file.
Ratify the org's consolidated stance on a claim document (ratify-gated: requires a role with ratify: true; propose-only roles are denied outright — there is no staged-proposal path for consolidation).
No parameter schema on file.
Record a tension — a contradiction or unresolved pull between two vault documents — to the advisory tension log.
No parameter schema on file.
Record the closure of a previously logged tension.
No parameter schema on file.
Compute connected components of the tension graph: groups of vault documents joined transitively by unresolved tensions.
No parameter schema on file.
Compute the transitive closure of downstream documents that cite or link a contested document — or the union over a contested cluster.
No parameter schema on file.
Compose the vault's LIVE tensions into a human-legible triage card: every unresolved, non-accepted tension grouped by cluster and annotated for a curator deciding what to resolve.
No parameter schema on file.
Run the advisory curation checks across the vault: stale files past TTL, orphan files with no inbound links or sources citations, old drafts, stagnant low-confidence files, deprecated files still linked from canonical ones, and questions raised but unanswered anywhere in the vault.
No parameter schema on file.
Return the write history of a single document from the provenance log: every create, update, append, promote, and deprecate recorded for it, oldest first.
No parameter schema on file.
Record a proposed change to the vault for later human ratification via vault_ratify.
No parameter schema on file.
Approve or reject a single pending staged action.
No parameter schema on file.
Record that a (re-)derivation observed a derives_from edge between two documents.
No parameter schema on file.
Record a case-2 contradiction: a re-derivation failed with no upstream change.
No parameter schema on file.
List derives_from edges with their live aged strength, strongest first.
No parameter schema on file.
Query the compiled consumes graph (#233): edges minted mechanically from run correlation — a write carrying a run_id consumes every document the same run read beforehand.
No parameter schema on file.
The reverse of the knowledge graph: given a target, list the documents that reference it.
No parameter schema on file.
Deterministic, LLM-free compatibility dispatch for a changed document (#232 tier 1).
No parameter schema on file.
The tier-2 semantic-review queue (#232): every declared/earned dependency whose upstream changed but whose compatibility the structure could not decide (tier 1's residual — #234's pending-unchecked class), minus pairs an existing verdict already covers.
No parameter schema on file.
Record the answer to a vault_tier2_queue item (#232 tier 2).
No parameter schema on file.
Edge staleness (#234): is a document stale WITH RESPECT TO its upstream inputs, and does it matter?
No parameter schema on file.
List all current board findings, grouped by disposition column (new / accepted / waiting / resolved / dismissed).
No parameter schema on file.
Record a human disposition decision on a board finding.
No parameter schema on file.
Deterministically verify that a finding no longer reproduces, and if so write a system-authored 'resolved' event.
No parameter schema on file.
- Every tool, no call limit
- No card, and mcprush charges nothing for it
- Source published under MIT
- Runs on your machine — nothing of it reaches our gateway
- Nothing to cap, because mcprush bills nothing
What counts against your monthly calls
| Tool | Unit | Calls used | Out of the allowance |
|---|
Nothing here is billable: mcprush charges nothing to install Daftari or to call it, at any volume.
Two independent axes, because powerful and malicious are different questions. The grade is threat only. The capability level is blast radius, and it is never a penalty on the grade — it is priced as one subtract-only term in the score, where you can see it.
| Term | Level | What it prices | Points |
|---|---|---|---|
| capability-exposure | high | capability blast radius (high) — client exposure if the model is manipulated | −6 |
| verification-discount | source | publisher verification (provenance) — cryptographic build provenance ties the artifact to its source | −0 |
| coverage-honesty | source | inspection depth (source) — how much of the target the scan could see | −0 |
What the scan could actually read
A grade is only as meaningful as its coverage, so the scanner publishes its own depth before it publishes its result.
No tools were enumerated, so prompt-injection, capability and toxic-flow analysis had no tool surface to inspect. To grade a package’s real runtime tools, scan the running server: --command "npx -y <package>".
dist/interview/index.d.ts (1 KB) was not read: over the file-count budget.
Capability — what it could do if the model were manipulated
Tags derived from each tool’s schema and the implementation, not from what the tool calls itself. high is the level these add up to.
| Tool | Capability tags | Why the tag was assigned |
|---|---|---|
| No per-tool rows for this release — see the note below. | ||
Toxic-flow graph
The lethal trifecta, checked as a graph rather than as a checklist: untrusted input, a sensitive source and an external sink have to meet before there is a path worth worrying about.
The public result for this release does not print the flow graph, so there is nothing to show here. That is not the same as "no paths were found": what the scan did read is above, under coverage.
Supply chain and provenance
This is the first scan of this surface here, so there is nothing yet to compare it against.
Every result on this tab comes from one deterministic pass over the published package — offline, rule by rule, and auditable line by line above. Same methodology version, same bytes, same score.
Release history
Always latest — the install command names no version, so it installs the newest release when it runs; the current release on file is 3.20.0. Pin a release below and the command asks for exactly that one.
No release note was published with this version.
No release note was published with this version.
- The version this listing was on when mcprush began following its releases. The registry was not asked when it shipped, so this row carries no date.
One review per account, from a signed-in account that does not publish this listing. Nothing else is asked — you can say what you think before you install it. Publishers can reply once per review, and a review can be edited or deleted by whoever wrote it.
Any signed-in account that does not publish Daftari can review it, once — before installing it or after. A review from an account that has installed it is marked as one.
Nobody has reviewed this listing. The rating on the card is the mean of the reviews written here and nothing else, so there is no rating until somebody writes the first — which takes a signed-in account that does not publish it, and nothing else.