Search arXiv papers, download full text, semantic search, citation graphs, and alerts via MCP.
Nothing to run from here
Papex 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.
96/100
trust score
170
downloads a day
local
runs on your machine
1k
downloads in 30d
What it does
Prepare private research drafts for your review in Papex and read public Papex records.
Quickstart
# 1 — add it to Claude Code
claude mcp add --scope user papex-mcp '--' npx -y @papexlabs/mcp
# 2 — ask your agent for something these tools do
# papex_seal_draft, papex_checkpoint, papex_restore_checkpoint and 20 more
Papex costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.
Publishernot claimed
PAAll 1 server
papex1 server
1 category
Collected from a public index. Nobody has claimed this account, so nothing here was written by its author — claim it if it is yours.
Installation
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 Papex 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.
Claude Codedeclared by this listing
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 papex-mcp '--' npx -y @papexlabs/mcpReconnect, or start a new session, and the tools appear in the model’s tool list.
Free on mcprush, and it runs from its publisher’s own package: this marketplace is not in the path, so there is no key of ours to hold and no card to add.
What it asks of you
Nothing is required, but it can take PAPEX_DELEGATE_PRIVATE_KEY (Delegate key from Papex → Workspace → Connections (shown once). Needed for any grant. Stays on this computer.), PAPEX_RESEARCH_DRAFT_GRANT_ID (Grant id that allows preparing private Hypothesis, Discovery and Publication drafts.), PAPEX_RESEARCH_DRAFT_ASSIST_GRANT_ID (Grant id that allows reading selected private-draft context and proposing changes.), PAPEX_RECORD_ARTIFACT_GRANT_ID (Grant id that allows proposing supporting artifacts into a private record version.), PAPEX_ARTIFACT_FILE_CUSTODY_GRANT_ID (Grant id that allows uploading an explicitly selected file set into private quarantine.), PAPEX_SEAL_GRANT_ID (Grant id that allows DELEGATE seals of selected local files or a draft fingerprint.), PAPEX_CHECKPOINT_GRANT_ID (Grant id that allows DELEGATE content checkpoints of a local directory.) if you want to set them.
What it writes
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.
What you approve
23 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.
On explicit user request, use a separate revocable seal grant to hash selected local files or commit an existing draft fingerprint. Delegated mode retains a local witness in .papex-seals (with a .gitignore so it stays out of version control) and sends only the root hash, count, kind, optional…
No parameter schema on file.
… Requires the separate checkpoint grant. Records a hidden content checkpoint: only the root hash of the directory's current file set is sent to Papex. The same bytes are also copied into a local content store outside that directory so the file set can be written out again later with…
No parameter schema on file.
Writes the files of one local checkpoint into a new or empty directory on this machine from the local content store, then re-hashes them and reports whether they reproduce the manifest hash that was signed. Reads nothing from Papex, touches no file in the watched directory, refuses a target inside…
No parameter schema on file.
Lists this machine's local checkpoint store as file then versions: content hash, size, local capture time, the checkpoints each version belongs to, and whether its content copy is still on this device. Each checkpoint also carries its id, manifest hash and file count, the same three values the…
No parameter schema on file.
Compares two local checkpoints, or one checkpoint with what its folder holds right now, on this machine only: which files were added, removed or changed, with sizes and, for text files up to 256 KiB on each side, a unified line diff (capped at … lines per file and … in total). Binary or larger…
No parameter schema on file.
Check whether this MCP server can prepare private Papex research drafts, provide bounded proposal-only draft assistance and reference lookups, prepare external contextual Artifact references, place private Artifact file packages in quarantine, seal a hash under a separate seal grant, or record…
No parameter schema on file.
No parameter schema on file.
No parameter schema on file.
Read and hash a local research directory on this machine, without uploading bytes or creating a receipt. Returns relative paths, local hashes and heuristic file roles to the connected AI app. Skipped: common repository, dependency and build folders, symlinks, files with more than one hard link,…
No parameter schema on file.
Create a PRIVATE hypothesis Draft on Papex from agent-prepared fields. Does NOT publish. Returns a review URL for human review and exact-version signing in the browser. Before calling, resolve or explicitly mark unresolved: system/population, variables, parameters, observer/reference frame,…
No parameter schema on file.
Create a PRIVATE, structured Discovery draft from a bounded result. Does NOT publish. Before calling, identify the problem addressed, result summary, method or derivation, evidence, limitations, references or basis records — the exact signed version (id, version, content hash) of a Papex…
No parameter schema on file.
Propose ONE contextual artifact — dataset, code, derivation, notebook, proof, protocol, model or supplement — against the EXACT version of one existing Papex record. The proposal is PRIVATE and stays PROPOSED until the accountable human resolves it in the browser; this tool cannot accept it,…
No parameter schema on file.
Hash and upload 1–32 explicitly selected Markdown, text, PDF, CSV, TSV, JSON, JSONL or notebook files into private Papex quarantine, bound to the exact version of an existing research record. The package and its parent link stay private and proposed. This tool cannot sign for the human, accept the…
No parameter schema on file.
No parameter schema on file.
Select an exact signed base (baseVersion and baseContentHash for private records), then hash and upload one replacement PDF into a base-bound private vN+1 review. The base may be private or public; its access is unchanged. Refuses to overwrite other private revision work. Optional claimScope…
No parameter schema on file.
Privately poll status of an agent-prepared research or Publication draft using its research-record-draft grant. Does not publish. Testable and Registered are workflow statuses, not scientific validation.
No parameter schema on file.
Privately poll one agent-prepared Discovery draft using its research-record-draft grant. Returns workflow state and the human review URL only. It cannot publish, expose human edits, or establish that the declared result is true or novel.
No parameter schema on file.
List only bounded summaries of the connected grant owner's private Hypothesis, Discovery, and Publication drafts so the human can select one. Returns no manuscript, affiliation, artifact decisions, fingerprint, signing data, or other user's drafts.
No parameter schema on file.
Read only human-selected, bounded authoring fields from one private draft owned by the connected grant. Affiliation, rights, Artifact decisions, lifecycle controls and signing fields are unavailable. This cannot save, accept, sign, or publish.
No parameter schema on file.
Create a proposal bound to one exact private draft version and fingerprint. The proposal does not mutate or save the draft. Only the accountable human can accept it, then explicitly save, decide affiliation and rights, resolve Artifacts, request signing, sign, or publish.
No parameter schema on file.
Ask Papex to look up one reference identifier in its registry: Crossref or DataCite for a DOI, arXiv, or PubMed. Returns whether it was found and, if so, the registered title, authors, year, venue (journal, proceedings or publisher) and link. The lookup only says the identifier exists and what it…
No parameter schema on file.
No parameter schema on file.
No parameter schema on file.
Free
$0 forever
- Every tool, no call limit
- No card, and mcprush charges nothing for it
- Source published under Apache-2.0
- 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 Papex or to call it, at any volume.
Call allowance · this installenforced at the gateway
Monthly allowance, in calls
—
—
Alert at 80% · then the tool returns a normal MCP error
Used this month: —
ATrust score 96 / 100
Release 0.8.1 · scanned 13 Oct 2026
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.
Threat score100
Client score96
Findings2 threat · 0 capability
Capabilitymoderate
Coveragesource
Verificationrepo
Surface digest6ea98b035970b8aaee8d65aaccaab720b4f322611cdede2f8d1a1776429f130d
Score vectorthreat 100 → client 96
| Term | Level | What it prices | Points |
|---|---|---|---|
| capability-exposure | moderate | capability blast radius (moderate) — client exposure if the model is manipulated | −3 |
| verification-discount | repo | publisher verification (public source) — no provenance, but the source is public and inspectable | −1 |
| coverage-honesty | source | inspection depth (source) — how much of the target the scan could see | −0 |
score = clamp(round(threatScore − E_cap − E_ver − E_cov), 0, 100). No model, no opinion, no network call at scoring time: the same methodology version against the same bytes always produces this number, and every line above is a field of the report rather than a summary of it.
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.
Tool surfaceread
Implementation sourceread
Package metadataread
Live transportnot available
Tools were statically extracted from the published source (23 recovered), not enumerated from a running server. Tool-poisoning, Unicode-smuggling, capability and toxic-flow analysis ran on this inferred surface, but a mis-parsed registration could be missed or mis-attributed, so tool-derived findings are capped below “confirmed”. To grade the real runtime surface, scan the running server: --command "npx -y <package>".
1 of the 23 recovered tools are registered only when t?.artifactFileCustodyGrantId holds (dist/index.js:92) — typically when a key or option is configured. They are part of the server's surface and were analysed; a server started without that configuration exposes only the other 20.
2 of the 23 recovered tools are registered only when dl(r,{config:i,api:e.api,account:e.account}),t?.researchDraftGrantId&&t.deriveHandoffSalt holds (dist/index.js:92) — typically when a key or option is configured. They are part of the server's surface and were analysed; a server started without that configuration exposes only the other 20.
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. moderate is the level these add up to.
20 tools are registered unconditionally · (1) 1 more only when t?.artifactFileCustodyGrantId holds (dist/index.js:92) · (2) 2 more only when dl(r,{config:i,api:e.api,account:e.account}),t?.researchDraftGrantId&&t.deriveHandoffSalt holds (dist/index.js:92) — typically when a key or option is configured. All 23 are in the table below and were analysed; the conditional ones are marked.
| Tool | Capability tags | Why the tag was assigned |
|---|---|---|
| papex_seal_draft | state-write | state-write: declares readOnlyHint: false |
| papex_checkpoint | state-write | state-write: declares readOnlyHint: false |
| papex_restore_checkpoint | state-write | state-write: declares readOnlyHint: false |
| papex_list_checkpoints | no tags | |
| papex_diff_checkpoints | state-write | state-write: declares readOnlyHint: false |
| papex_setup_delegation | untrusted-input | untrusted-input: keyword "download" |
| papex_get_profile | no tags | |
| papex_list_records | no tags | |
| papex_preview_capsule | state-write | state-write: declares readOnlyHint: false |
| papex_prepare_hypothesis | state-write | state-write: declares readOnlyHint: false |
| papex_prepare_discovery | state-write | state-write: declares readOnlyHint: false |
| papex_propose_record_artifact | state-write | state-write: declares readOnlyHint: false |
| papex_upload_artifact_package conditional (1) | state-write | state-write: declares readOnlyHint: false |
| papex_prepare_publication conditional (2) | state-write | state-write: declares readOnlyHint: false |
| papex_prepare_publication_revision conditional (2) | state-write | state-write: declares readOnlyHint: false |
| papex_get_publication_status | no tags | |
| papex_get_discovery_status | no tags | |
| papex_list_private_drafts | no tags | |
| papex_get_draft_assist_context | no tags | |
| papex_propose_draft_changes | state-write | state-write: declares readOnlyHint: false |
| papex_lookup_reference | untrusted-input | untrusted-input: declares openWorldHint: true (interacts with entities outside the server) |
| papex_discard_draft | state-write | state-write: declares readOnlyHint: false, destructiveHint: true |
| papex_session_checkpoint | state-write | state-write: declares readOnlyHint: false |
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
Integrityfirst-seen
Surface digest6ea98b035970b8aa
Scanned2026-10-13
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 0.8.1. Pin a release below and the command asks for exactly that one.
0.8.1
12 Oct 2026
latestNo release note was published with this version.
0.8.0
8 Oct 2026
No release note was published with this version.
0.7.3
undated
- 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.
—
no reviews yet
Who can review
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.
Sign in to write one
Any signed-in account that does not publish Papex can review it, once — before installing it or after. A review from an account that has installed it is marked as one.
No reviews yet
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.