Noeis
NPM · @NOEIS/WIKI-MCP
Your reading, cited wiki and agent-kept editions in Noeis. Read-only unless you grant more.
Available components
Last scanned (17 hours ago)
Scores higher than 38% of npm servers
Category Productivity and Notes
How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →
Supply Chain Security81
- No malware found by supply-chain analysis.Pass
- CVE check failed: a known high-severity CVE affects zod 3.25.76, a direct dependency. A fixed version is available. View diagnostics → Fail
- No install/post-install scripts declared.Pass
- 31 of 93 dependencies flagged as unhealthy. View diagnostics → Partial
Provenance & Transparency45
- Source repository is publicly reachable at the declared URL. View diagnostics → Pass
- Provenance check failed: no build-provenance attestation is published. See how to fix → View diagnostics → Fail
- Clear OSI-approved license (ISC).Pass
- Actively maintained (last published 5 days ago).Pass
- Disclosure check failed: no security disclosure policy was found in the source repository. See how to fix → Fail
Schema Quality & AI Usability75
- 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).Pass
- AI-judged instruction clarity (good).Pass
- Context-footprint check failed: tool/resource definitions use about 1944 tokens (~162/item across 12 items; 12 tools + 0 resources), over budget; trim descriptions and params. See how to fix → Fail
- Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management20
- Stability observed for 6 of 30 days with no destabilising changes; credit accrues until the full window elapses.Partial
Tool Coverage87
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 62% of tool parameters carry a description.Partial
Tool Safety100
- No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
- We read all 12 captured tool definition(s), and no name or description among them implies an irreversible operation.Pass
- An AI judge read all 13 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
- Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.Pass
Checks we ran (16)
- Malware scan
- Known CVEs
- Install scripts
- Dependency health
- Source repository
- Provenance
- License
- Maintenance
- Security disclosure
- Schema quality
- Stability
- Tool coverage
- Injection markers
- Destructive annotations
- Judged manipulation
- MCP protocol
Does not apply to this package (2)
- Endpoint SecurityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.
- Transport & ReachabilityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.
The score does not measure everything. What the score does not measure →
Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.
- 21 Nov 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 17 to 20. That category is still filling its 30-day observation window: 5 days of observed history at the previous scan, 6 at this one. The score rises as the window fills, whether or not the server changes.
- 19 Nov 26 +2
- Stability: unverified → 0.13 ▲ functional
- 16 Nov 26 +21
- CVE-2023-54404 affects this package: high ▼ security
- Known CVEs: unverified → fail ▼ security
- Malware scan: unverified → pass ▲ security
- Dependency health: unverified → 0.85 ▲ functional
- 15 Nov 26 43
First indexed and scored.
Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.
Captured 21 Nov 2026 · Analysed npm/@noeis/wiki-mcp@0.5.1
Provenance No attestation
The registry publishes no build provenance for this version, so there is nothing to verify.
| Result | No attestation |
|---|---|
| Ecosystem | npm |
Background: How many MCP packages publish verified provenance →
Vulnerabilities 1 finding
| ID | CVE | Severity | Vector | Fix available |
|---|---|---|---|---|
| GHSA-mrhx-c7xc-m3w5 | CVE-2023-54404 | high | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H | yes |
Background: What a vulnerability scan can and cannot prove →
Dependencies 93 packages
| Packages resolved | 93 |
|---|---|
| Stale | 31 |
| Tree resolution | Complete |
Background: SBOMs and build attestations, explained →
The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →
connection_info ~36
Verify this exact NOEIS grant with a metadata-only authenticated read. Returns stable workspace identity and enforced scopes without reading or creating user content.
Input schema present but exposes no named parameters.
No output schema declared.
No examples provided.
create_article ~274
THE way an article gets into the reader's Library, where they open and read it. Every "save this", "add this to my library", "keep this article" is this tool — ingest_source is the wiki and leaves the Library untouched. Pass the full article text as content: without it the Library holds a title and a link, and the reader shows a highlight-only edition instead of the piece. Read the page first and send what you read; the server only falls back to fetching the URL itself when content is missing, and a paywall or a script-rendered page defeats it. The result says which of the three happened. Save what you read for the reader as you go: it lands in pencil, on their "Read by agents" shelf, until they ink it. placement later puts it on their nightstand to read themselves.
| Name | Type | Req | Description |
|---|---|---|---|
| author | string | – | – |
| content | string | – | The full article text or HTML. Omit only when the body is genuinely unavailable to you. |
| folderId | string | – | – |
| placement | string | – | later: the nightstand, for the reader to read themselves. Omit for the Library. |
| publicationDate | string | – | – |
| siteName | string | – | – |
| title | string | yes | – |
| url | string | yes | – |
No output schema declared.
No examples provided.
create_highlight ~116
Create a highlight on an existing Library article using an exact source passage, never a paraphrase or invented quotation. The note is the reader’s private thought, kept separate from source text. Save a thought only when the reader explicitly asks; preserve their words and label any assistant wording. Read the source first.
| Name | Type | Req | Description |
|---|---|---|---|
| anchor | object | – | – |
| articleId | string | yes | Library article id. |
| color | string | – | – |
| note | string | – | – |
| tags | array | – | – |
| text | string | yes | – |
No output schema declared.
No examples provided.
create_notebook_entry ~216
Write a new Notebook entry. A note is the reader's own thinking, not a saved source — save sources with create_article. Name a folder by folderId or folder (the name, resolved for you). Kind is claim, evidence or note; evidence may name the claim entry it supports.
| Name | Type | Req | Description |
|---|---|---|---|
| blocks | array | – | Structured blocks, when you have them. Replaces content. |
| claimId | string | – | The claim entry this evidence supports. Only holds when type is evidence. |
| content | string | – | The note's text. Becomes a single paragraph block when blocks are not given. |
| folder | string | – | Notebook folder name, resolved case-insensitively. Ignored when folderId is given. |
| folderId | string | – | Notebook folder id. |
| linkedArticleId | string | – | A Library article this note is about. |
| tags | array | – | – |
| title | string | yes | The note's title. |
| type | string | – | What the note is. Defaults to note. |
No output schema declared.
No examples provided.
create_page ~164
Create a new private draft wiki page. Use this only after checking search_pages/list_pages to avoid duplicates. Requires an agent-write token.
| Name | Type | Req | Description |
|---|---|---|---|
| body | – | – | Optional body as plain text or TipTap JSON. Plain text keeps its paragraph breaks; blank lines separate paragraphs. |
| initialSourceRef | object | – | Optional initial source reference to attach, e.g. { type: "external", url, title }. Do not send sourceRefs, claims, or citations — the API rejects them; those ledgers are managed by wiki maintenance. |
| initialSourceRefs | array | – | Optional initial source references when a page cites more than one, up to 8. |
| pageType | string | – | – |
| sourceScope | string | – | – |
| title | string | yes | Page title. |
No output schema declared.
No examples provided.
file_edition_items ~416
Add what you found today to the issue this moment belongs to, WITHOUT resending what is already there. Noeis picks the issue from the topic's cadence, so two agents filing the same day file into the same one. Existing research items and reader saves remain. If another agent already filed the same link, your item is kept as a second reading beside theirs (your own repeats are skipped). Write your own finding and boundary; do not paraphrase theirs. Every item needs its boundary. If you looked at a section and nothing met your bar, say so in `checked`. A section you leave silent reads to the reader as 'not reported'. A supplied nonempty `headline` replaces the issue's current headline. A `followUps` answer for an already answered watch line replaces its status and note. These replacements keep no history and have no undo through this connection; read the current issue and confirm the intended replacement with the reader first. Adding new items can also update the issue's masthead attribution. This tool cannot delete research items or publish the issue. Send items, checked, followUps or a headline. Requires an agent-write token.
| Name | Type | Req | Description |
|---|---|---|---|
| checked | array | – | Sections you looked at where nothing met your bar: a key, or { section, note }. A section you leave silent reads to the reader as "not reported". |
| followUps | array | – | What became of the last issue's watch list. Each names a line it printed; one it never printed is refused. |
| headline | string | – | What the week comes to, in one plain sentence a reader would say: the takeaway, not a topic. |
| items | array | – | Only the new findings. What is already filed stays. May be left out when you send checked. |
| profile | string | yes | Topic key, from list_edition_profiles. |
| standfirst | string | – | Standfirst, used only when this opens a new issue. |
| title | string | – | Title, used only when this opens a new issue. |
No output schema declared.
No examples provided.
get_page_markdown ~32
Get a wiki page rendered as Markdown.
| Name | Type | Req | Description |
|---|---|---|---|
| pageId | string | yes | Noeis wiki page id. |
No output schema declared.
No examples provided.
list_edition_profiles ~57
List the reader's edition topics: what each paper is called, its sections, and how often it comes out (daily, weekly, monthly). Call this before filing so you use a topic that exists and a section it actually has.
Input schema present but exposes no named parameters.
No output schema declared.
No examples provided.
list_pages ~74
Find candidate wiki pages to inspect. Returns lightweight page rows with id, title, pageType, slug, and updatedAt.
| Name | Type | Req | Description |
|---|---|---|---|
| limit | number | – | – |
| pageType | string | – | – |
| q | string | – | Optional title/body search query. |
| status | string | – | – |
| visibility | string | – | – |
No output schema declared.
No examples provided.
read ~58
Read one thing recall found, in full: an article with all its highlights and margin notes, a Think note, or a wiki page. Pass the `id` of a recall hit.
| Name | Type | Req | Description |
|---|---|---|---|
| id | string | yes | The id of a recall hit. |
No output schema declared.
No examples provided.
recall ~164
Ask everything the reader keeps: saved articles and their highlights with margin notes, Think notes, concepts, the views they hold, wiki pages and filed Editions. Ask in plain words, the way you would ask the reader; results are the passages that answer it, best first, each with its source id and a link. Start here for any question about what the reader has read, written or believes. Open a hit with the tool in its `open` field. A hit with `pencil` was written by the agent it names and not yet taken up by the reader: never present it as their view. Nothing returned means nothing they keep bears on it.
| Name | Type | Req | Description |
|---|---|---|---|
| limit | integer | – | Most passages to return. |
| query | string | yes | The question, in plain words. |
No output schema declared.
No examples provided.
update_notebook_entry ~167
Change a Notebook entry. Only the fields you pass are touched, so an update that renames a note leaves its text, tags and filing standing.
| Name | Type | Req | Description |
|---|---|---|---|
| blocks | array | – | Replaces the note's blocks. |
| claimId | string | – | Only holds when type is evidence. |
| content | string | – | Replaces the note's text and its blocks. |
| entryId | string | yes | Notebook entry id. |
| folder | string | – | Notebook folder name. Ignored when folderId is given. |
| folderId | string | – | Notebook folder id. Pass an empty folder name to unfile. |
| linkedArticleId | string | – | – |
| tags | array | – | Replaces the existing tags. |
| title | string | – | – |
| type | string | – | – |
No output schema declared.
No examples provided.
How do I install the Noeis MCP server?
Noeis runs locally as an npm package, launched with npx -y @noeis/wiki-mcp. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
npm · @noeis/wiki-mcp
claude mcp add atsokolas-noeis -- npx -y @noeis/wiki-mcp
What is the Noeis MCP server?
Noeis is an MCP server listed in the public MCP registry as io.github.atsokolas/noeis. Your reading, cited wiki and agent-kept editions in Noeis. Read-only unless you grant more. This page covers its npm package (@noeis/wiki-mcp).
Is the Noeis MCP server safe to use?
Noeis scores 67 out of 100 on VerifyMCP. We recorded 1 known advisory against it as of 21 November 2026. It declares no install or post-install scripts. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.
What tools does the Noeis MCP server expose?
Noeis exposes 12 tools: connection_info, recall, read, list_edition_profiles, list_pages, and 7 more. Their descriptions and schemas cost roughly 1,774 tokens of context every time the server is loaded.
Is the Noeis MCP server still maintained?
Noeis is still listed as active in the MCP registry. We last reached this channel on 21 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.
What licence is the Noeis MCP server under?
Noeis declares the ISC licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.
What did VerifyMCP test on the Noeis MCP server?
VerifyMCP ran 16 checks on Noeis, covering Malware scan, Known CVEs, Install scripts, Dependency health and 12 more. Endpoint Security and Transport & Reachability do not apply to an npm package, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.