# passwerk (npm · @passwerk/server)

Offline EU Battery Passport compiler: ingest documents, map, validate (IDTA 02035), emit AASX.

- Trust score: 92/100 (high trust)
- Change this week: +3
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-10-17

## Components

- npm · `@passwerk/server`: 92/100 (this document), [markdown](https://verifymcp.io/servers/shahriarbijoy-passwerk/passwerk-server.md), [page](https://verifymcp.io/servers/shahriarbijoy-passwerk/passwerk-server)

## Channel facts

- Registry: `npm`
- Package: `@passwerk/server`
- Version: `0.1.1`
- Transport: `stdio`

## Trust breakdown

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. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-10-17.

- **Supply Chain Security**: 81/100
  - No malware found by supply-chain analysis.
  - CVE check failed: a known high-severity CVE affects @modelcontextprotocol/sdk 1.30.0, a direct dependency. A fixed version is available.
  - No install/post-install scripts declared.
  - 31 of 120 dependencies flagged as unhealthy.
- **Provenance & Transparency**: 100/100
  - Source repository is publicly reachable at the declared URL.
  - Cryptographically verified build provenance (signed, bound to ShahriarBijoy/passwerk).
  - Clear OSI-approved license (Apache-2.0).
  - Actively maintained (last published 33 days ago).
  - Publishes a security disclosure policy (SECURITY.md).
- **Schema Quality & AI Usability**: 88/100
  - 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).
  - AI-judged instruction clarity (excellent).
  - Tool/resource definitions use about 2810 tokens (~87/item across 32 items; 12 tools + 20 resources), lean.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 100/100
  - No destabilizing schema changes in the last 30 days.
- **Tool Coverage**: 100/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 100% of tool parameters carry a description.
  - Structured output schemas are declared (100% of tools); any adoption earns full credit.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - We read all 12 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 13 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.
  - Supports UI / widget rendering.

## Install

### How do I install the passwerk MCP server?

passwerk runs locally as an npm package, launched with npx -y @passwerk/server. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add shahriarbijoy-passwerk -- npx -y @passwerk/server
```

### Cursor

```json
{
  "mcpServers": {
    "shahriarbijoy-passwerk": {
      "command": "npx",
      "args": [
        "-y",
        "@passwerk/server"
      ]
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "shahriarbijoy-passwerk": {
      "command": "npx",
      "args": [
        "-y",
        "@passwerk/server"
      ]
    }
  }
}
```

### Codex

```bash
codex mcp add shahriarbijoy-passwerk -- npx -y @passwerk/server
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "shahriarbijoy-passwerk": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "@passwerk/server"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add shahriarbijoy-passwerk --command npx --arg -y --arg @passwerk/server
```

### Hermes

```yaml
mcp_servers:
  shahriarbijoy-passwerk:
    command: "npx"
    args: ["-y", "@passwerk/server"]
```

### Netclaw

```json
{
  "McpServers": {
    "shahriarbijoy-passwerk": {
      "Transport": "stdio",
      "Command": "npx",
      "Arguments": [
        "-y",
        "@passwerk/server"
      ]
    }
  }
}
```

### Vellum

```bash
assistant mcp add shahriarbijoy-passwerk -t stdio -c npx -a -y @passwerk/server
```

### Other

```json
{
  "mcpServers": {
    "shahriarbijoy-passwerk": {
      "command": "npx",
      "args": [
        "-y",
        "@passwerk/server"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-10-17 (score 92, +2)

- [security] Stability: 0.83 → pass

### 2026-10-16 (score 90, +1)

No change was recorded against any check on this day. Stability & Change Management went from 80 to 83. That category is still filling its 30-day observation window: 24 days of observed history at the previous scan, 25 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-10-14 (score 89, −3)

- [functional] Stability: pass → 0.77

### 2026-10-13 (score 92, 0)

- [security] Stability: 0.97 → pass

### 2026-10-12 (score 92, +1)

- [functional improvement] Security disclosure: unverified → pass

### 2026-10-11 (score 91, +2)

- [functional regression] Security disclosure: pass → unverified

### 2026-10-07 (score 89, −4)

- [security regression] CVE-2026-104850 affects this package: high
- [security regression] Known CVEs: pass → fail

### 2026-10-03 (score 93, +1)

No change was recorded against any check on this day. Stability & Change Management went from 63 to 67. That category is still filling its 30-day observation window: 19 days of observed history at the previous scan, 20 at this one. The score rises as the window fills, whether or not the server changes.

## MCP tools (12)

### `ingest_documents` (~203 tokens)

Ingest supplier documents

Reads PDF, XLSX, CSV, DOCX and TXT documents into a DocumentBundle with page, line, table and cell provenance, deterministically and without OCR or a model. Returns a per-document summary and a bundleId to pass to extract_facts; ask for detail "full" to see the pages. Unsupported or unreadable files are reported, not dropped.

Input parameters:

- `detail` (string): summary (default): per-document counts and the bundleId; full: the whole DocumentBundle
- `inline` (array): Document bytes as base64, for hosts that cannot pass paths
- `lang` (string): Language of the text summary (default en)
- `limits` (object): Partial IngestLimits override (maxInputBytes, maxCells, ...)
- `paths` (array): Files or directories on the server’s file system (a directory contributes its PDF, XLSX, CSV, DOCX and TXT files, non-recursive). Preferred for local installs

Output parameters:

- `bundle` (object)
- `bundleId` (string)
- `documents` (array)
- `error` (string): Set when the call failed
- `errors` (array)

### `extract_facts` (~98 tokens)

Extract facts

Turns a DocumentBundle into a FactSet: labelled values with normalised numbers, units and dates, each with file, page, line or cell provenance. Pure heuristics, no model. Returns the FactSet and a factSetId to pass to suggest_mappings.

Input parameters:

- `bundle` (required): A DocumentBundle object, or { bundleId } as returned by ingest_documents
- `lang` (string): Language of the text summary (default en)

Output parameters:

- `bundleId` (string)
- `error` (string): Set when the call failed
- `factSetId` (string)
- `facts` (object)

### `suggest_mappings` (~268 tokens)

Suggest attribute mappings

Scores every fact against the knowledge-base synonym index (DE/EN), unit and type checks and the attribute range, and returns candidate mappings sorted by confidence with a DE/EN explanation and provenance. Review proposals under 0.7 with the user or the source page, then pass the accepted ones to apply_mappings. Ask for detail "full", or pass attributeIds, to see every candidate with its file, page and cell in the text.

Input parameters:

- `attributeIds` (array): Keep only proposals for these attribute ids; unknown ids are reported, not an error. When given, the text lists every match, not just the first 10, with provenance, even in summary mode
- `category` (string, required): Passport category; decides applicability
- `detail` (string): summary (default): header line plus the first 10 proposals, no provenance. full: every proposal after the filters, one line each, with provenance (file, page, cell); no cap
- `facts` (required): A FactSet object, or { factSetId } as returned by extract_facts
- `lang` (string): Language of the text summary (default en)
- `minConfidence` (number): Drop proposals below this confidence (default 0). 0.7 is the review threshold

Output parameters:

- `category` (string)
- `counts` (object)
- `error` (string): Set when the call failed
- `factSetId` (string)
- `proposals` (array)

### `apply_mappings` (~141 tokens)

Apply mapping decisions

Folds accepted mapping decisions (or manual values) into a PassportDraft. Idempotent: the same decision twice is a no-op; a different value for an attribute that already has one is reported as a conflict and not applied unless override is true. Pass meta instead of draft to start a new draft.

Input parameters:

- `draft`: The draft to extend. Omit and pass meta to start a new one
- `lang` (string): Language of the text summary (default en)
- `mappings` (array, required): Accepted proposals from suggest_mappings, or manual values
- `meta` (object): Starts a new draft. Exactly one of draft or meta

Output parameters:

- `applied` (number)
- `conflicts` (array)
- `draft` (object)
- `draftId` (string)
- `error` (string): Set when the call failed

### `validate_passport` (~177 tokens)

Validate a passport draft

Runs all four validation layers on a PassportDraft: L1 schema, L2 AAS meta-model (aas-core), L3 IDTA 02035 template conformance, L4 domain plausibility. Returns the verdict (valid, valid_with_warnings, invalid) and every finding with layer, rule id, path, DE/EN message, legal reference and fix hint. The only source of a validity claim.

Input parameters:

- `asOf` (string): ISO date-time treated as "now" by the plausibility layer; default: draft.meta.createdAt
- `draft` (required): A PassportDraft object, or { draftId } as returned by an earlier call
- `lang` (string): Language of the text summary (default en)
- `skipPlausibility` (boolean): Skip L4 (domain plausibility). Default false

Output parameters:

- `draftId` (string)
- `error` (string): Set when the call failed
- `findings` (array)
- `layers` (object)
- `verdict` (string)

### `gap_report` (~270 tokens)

Gap report

The to-do list for a PassportDraft: every attribute bucketed as required, conditional, deferred or optional with status, legal references, who typically has the data and a suggested action; completeness for mandatory and overall data points, grouped by submodel and by data owner. Runs validate_passport first so invalid values count as gaps. Ask for detail "full" and a status/bucket filter to see every open attribute with its legal reference and next action, unabridged, for use without the structured content. Not legal advice.

Input parameters:

- `asOf` (string): ISO date-time treated as "now" for applicability; default: draft.meta.createdAt
- `bucket` (array): Keep only items in one of these buckets; completeness stays unfiltered
- `detail` (string): summary (default): open mandatory items, capped at 10. full: every item after the status/bucket filters, grouped by data owner, one line each with status, bucket, legal refs and the suggested next ac…
- `draft` (required): A PassportDraft object, or { draftId } as returned by an earlier call
- `lang` (string): Language of the text summary (default en)
- `status` (array): Keep only items with one of these statuses; completeness stays unfiltered

Output parameters:

- `byDataOwner` (array)
- `bySubmodel` (array)
- `category` (string)
- `completeness` (object)
- `draftId` (string)
- `error` (string): Set when the call failed
- `filter` (object)
- `isNotLegalAdvice` (boolean)
- `items` (array)
- `sources` (array)

### `review_passport` (~146 tokens)

Review a passport in the workbench

Opens the interactive passwerk workbench in hosts that render MCP Apps (Claude Desktop, Claude web): upload documents, accept or edit mapping proposals, read the gap report and export, all on the user’s machine. Pass the current draft so the user reviews it visually; the workbench keeps the draft id in sync so gap_report and emit_passport can continue on it. Hosts without a UI receive the same draft, validation report and gap report as text.

Input parameters:

- `draft`: The draft to review; omit to open an empty workbench
- `facts`: Extracted facts to review alongside the draft
- `lang` (string): Language of the text summary (default en)

Output parameters:

- `draft` (object)
- `draftId` (string)
- `error` (string): Set when the call failed
- `factSetId` (string)
- `facts` (object)
- `gap` (object)
- `report` (object)

### `emit_passport` (~244 tokens)

Emit the passport files

Emits a PassportDraft as AAS JSON (IDTA 02035 submodels), an AASX package, the draft JSON and/or the HTML sheet, and re-validates the emitted output. Fail-honest: files are returned even when the verdict is invalid, and the verdict is never better than validate_passport’s. Bytes come back base64 inline unless outDir is given.

Input parameters:

- `asOf` (string): ISO date-time treated as "now" by the re-validation
- `draft` (required): A PassportDraft object, or { draftId } as returned by an earlier call
- `htmlLang` (string): Language the html sheet opens in (both are in the file; default en)
- `lang` (string): Language of the text summary (default en)
- `outDir` (string): Directory to write into (needs a file system). Omit to receive base64 bytes inline
- `targets` (array, required): aas-json: AAS v3 JSON environment (IDTA 02035); aasx: AASX package; draft-json: the neutral PassportDraft; html: self-contained HTML passport sheet (DE and EN inside)

Output parameters:

- `draftId` (string)
- `error` (string): Set when the call failed
- `files` (array)
- `findings` (array)
- `verdict` (string)

### `generate_carrier` (~238 tokens)

Generate the data carrier (QR code)

Builds the data carrier for a passport: the unique identifier (the draft’s passportId, an absolute https URI), an optional GS1 Digital Link (/01/{gtin}/21/{serial} or /8004/{giai}) and the QR code as SVG or PNG. The QR encodes the Digital Link when GS1 data is given, else the identifier. Says nothing about validity; validate_passport does.

Input parameters:

- `draft`: The draft whose meta.passportId is the identifier
- `format` (string): Image format (default svg)
- `gs1`: GS1 key: GTIN (8, 12, 13 or 14 digits) plus serial, or a GIAI
- `lang` (string): Language of the text summary (default en)
- `outDir` (string): Directory to write the image into (needs a file system). Omit to receive base64 bytes inline
- `resolverBase` (string): https base of the GS1 Digital Link resolver; required with gs1
- `uid` (string): The passport identifier (absolute https URI) when no draft is given

Output parameters:

- `digitalLink` (string)
- `draftId` (string)
- `error` (string): Set when the call failed
- `format` (string)
- `image` (object)
- `isNotLegalAdvice` (boolean)
- `mediaType` (string)
- `payload` (string)
- `sources` (array)
- `uid` (string)

### `check_obligations` (~198 tokens)

Check battery passport obligations

Whether a battery passport is required for a battery type, role, energy and market date under Regulation (EU) 2023/1542 Article 77(1); returns the category, the mandatory, conditional and deferred attribute sets and the legal timeline. Answers required, not_required or insufficient_input. Not legal advice.

Input parameters:

- `asOf` (string): ISO date treated as "now"; default: the server clock
- `batteryType` (string, required): Battery type as placed on the market
- `energyKwh` (string): Battery energy in kWh as a decimal string, e.g. "2.5". Decides the industrial 2 kWh threshold
- `lang` (string): Language of the text summary (default en)
- `placedOnMarketDate` (string): ISO date the battery is or was placed on the market or put into service
- `role` (string, required): Economic operator role of the asker

Output parameters:

- `asOf` (string)
- `category` (string|null)
- `conditionalAttributes` (array)
- `error` (string): Set when the call failed
- `isNotLegalAdvice` (boolean)
- `mandatoryAttributes` (array)
- `missingInput` (array)
- `reason` (object)
- `sources` (array)
- `verdict` (string)

### `explain_attribute` (~126 tokens)

Explain an attribute or rule

Everything the knowledge base holds about one passport attribute (official DIN definition, DE/EN explanation, synonyms, who typically has the data, unit, range, applicability per category, legal references, IDTA template elements) or one plausibility rule (message, fix hint, legal reference). Nothing is composed; unknown ids are an error.

Input parameters:

- `id` (string, required): A knowledge-base attribute id (e.g. "batteryChemistry") or a rule id (e.g. "PW-PLAUS-001")
- `lang` (string): Language of the text summary (default en)

Output parameters:

- `error` (string): Set when the call failed
- `id` (string)
- `isNotLegalAdvice` (boolean)
- `kind` (string)
- `legalRef` (string|null)
- `legalRefs` (array)

### `list_capabilities` (~74 tokens)

List capabilities

Versions of the bundled IDTA 02035 templates and AAS schemas, knowledge-base statistics, when the artefacts were retrieved, the sovereignty statement, and this server’s tools and session-store usage. Call first to learn what this server can do.

Input parameters:

- `lang` (string): Language of the text summary (default en)

Output parameters:

- `aasSchemaVersions` (array)
- `error` (string): Set when the call failed
- `knowledgeBase` (object)
- `package` (string)
- `server` (object)
- `sovereignty` (string)
- `templates` (array)

## Diagnostics

Captured diagnostic sections: Provenance, Vulnerabilities, Dependencies. The full working is on the page: https://verifymcp.io/servers/shahriarbijoy-passwerk/passwerk-server#diagnostics

## Score history

- 2026-10-17: 92
- 2026-10-16: 90
- 2026-10-15: 89
- 2026-10-14: 89
- 2026-10-13: 92
- 2026-10-12: 92
- 2026-10-11: 91
- 2026-10-07: 89
- 2026-10-04: 93
- 2026-10-03: 93
- 2026-10-02: 92
- 2026-10-01: 92
- 2026-09-30: 92
- 2026-09-29: 91
- 2026-09-28: 91
- 2026-09-27: 90
- 2026-09-26: 90
- 2026-09-25: 89
- 2026-09-24: 89
- 2026-09-23: 88
- 2026-09-22: 88
- 2026-09-21: 84
- 2026-09-20: 84
- 2026-09-19: 84
- 2026-09-18: 84
- 2026-09-17: 84
- 2026-09-16: 84
- 2026-09-15: 84
- 2026-09-14: 84
- 2026-09-13: 69

## Common questions

### What is the passwerk MCP server?

passwerk is an MCP server listed in the public MCP registry as io.github.ShahriarBijoy/passwerk. Offline EU Battery Passport compiler: ingest documents, map, validate (IDTA 02035), emit AASX. This page covers its npm package (@passwerk/server).

### Is the passwerk MCP server safe to use?

passwerk scores 92 out of 100 on VerifyMCP. We recorded 1 known advisory against it as of 17 October 2026. It declares no install or post-install scripts. Its build provenance is signed and verified. 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 passwerk MCP server expose?

passwerk exposes 12 tools: ingest_documents, extract_facts, suggest_mappings, apply_mappings, validate_passport, and 7 more. Their descriptions and schemas cost roughly 2,183 tokens of context every time the server is loaded.

### Is the passwerk MCP server still maintained?

passwerk is still listed as active in the MCP registry. We last reached this channel on 17 October 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 passwerk MCP server under?

passwerk declares the Apache-2.0 licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

## Links

- npm package: https://www.npmjs.com/package/@passwerk/server
- Socket report: https://socket.dev/npm/package/@passwerk/server
- Repository: https://github.com/ShahriarBijoy/passwerk
- Changelog RSS feed: https://verifymcp.io/servers/shahriarbijoy-passwerk/passwerk-server.xml
- Changelog JSON feed: https://verifymcp.io/servers/shahriarbijoy-passwerk/passwerk-server.json
- HTML version of this page: https://verifymcp.io/servers/shahriarbijoy-passwerk/passwerk-server