DocuSense
PYPI · DOCUSENSE-MCP
Agents read PDFs and docs by section and claim: ~20x fewer tokens, hash-verified citations, diffs.
Available components
Last scanned (14 hours ago)
Scores higher than 58% of PyPI servers
Category Documents and Files
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 Security49
- Malware scan not yet available for this package.Unverified
- Known CVEs were checked across the 84 of 86 dependencies we could resolve, so this covers what we could see, not the whole tree.Partial
- Runs uv_build at install time, a recognised build step with no custom scripting around it. View diagnostics → Pass
- Dependency health was assessed across the 84 of 86 dependencies we could resolve, so this covers what we could see, not the whole tree. View diagnostics → Partial
Provenance & Transparency97
- Source repository is publicly reachable at the declared URL. View diagnostics → Pass
- Cryptographically verified build provenance (signed, bound to hksharma2004/docu-sense). View diagnostics → Pass
- Clear OSI-approved license (MIT).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 Usability76
- 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
- Tool/resource definitions use about 1735 tokens (~96/item across 18 items; 17 tools + 1 resources), lean.Pass
- Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management17
- Stability observed for 5 of 30 days with no destabilising changes; credit accrues until the full window elapses.Partial
Tool Coverage67
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 0% of tool parameters carry a description.Fail
Tool Safety100
- No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
- We read all 17 captured tool definition(s), and no name or description among them implies an irreversible operation.Pass
- An AI judge read all 19 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
- Implements a current MCP spec version (2026-07-28).Pass
Checks we ran (15)
- 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
Checks we could not complete (1)
These are gaps on our side, not findings about this package. We only credit what we can confirm, so a category made up entirely of them scores 0.
- Malware scan in Supply Chain SecurityMalware scan not yet available for this package.
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 −14
- Malware scan: pass → unverified ▼ security
- Security disclosure: unverified → fail ▼ functional
- 20 Nov 26 +6
- Source repository: fail → pass ▲ security
- 19 Nov 26 +16
- Malware scan: unverified → pass ▲ security
- Stability: unverified → 0.10 ▲ functional
- 16 Nov 26 58
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 pypi/docusense-mcp@0.1.2
Provenance Verified
A signed build attestation was found and verified, binding this exact artifact to the source repository it claims to come from.
| Result | Verified |
|---|---|
| Ecosystem | pypi |
| Reason | Verified |
| Discovered via | Registry attestation endpoint |
| Source repo | hksharma2004/docu-sense |
| Certificate issuer | https://token.actions.githubusercontent.com |
| Certificate SAN | https://github.com/hksharma2004/docu-sense/.github/workflows/release.yml@refs/tags/v0.1.2 |
| Rekor log index | 3918983702 |
| Predicate type | PyPI publish attestation https://docs.pypi.org/attestations/publish/v1 |
| Subject digest | sha256:026aca35310d75aa00ec2d8906583ba22bd1b26e327023c6a536b8d3ce7009b6 |
Background: How many MCP packages publish verified provenance →
Install scripts 1 script
| Hook | Tier | Command |
|---|---|---|
| build_backend | Known build tool | uv_build |
Background: Why install scripts are a supply-chain risk →
Dependencies 84 packages
| Packages resolved | 84 |
|---|---|
| Stale | 9 |
| No linked repository | 4 |
| Tree resolution | Partial |
The dependency tree was only partially resolved, so these counts may be incomplete.
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 →
add_document ~155
Add a local file or folder (.pdf, .docx, .html, .md, .txt) to DocuSense. Re-adding a file whose content changed appends a new version (v1, v2, …); unchanged files are skipped, so it is safe to call repeatedly. doc / label set the document id / version label (single file only). Ingestion runs in the background: if it is not finished within wait_seconds the result has status 'running' and a job_id; call ingest_status(job_id) until it is 'done'.
| Name | Type | Req | Description |
|---|---|---|---|
| doc | – | – | – |
| label | – | – | – |
| path | string | yes | – |
| title | – | – | – |
| wait_seconds | number | – | – |
No output schema declared.
No examples provided.
assemble_context ~87
Build the final evidence pack for an answer. For each span: verify its SHA-256 (rejecting tampered spans), attach the latest version's text if the span is obsolete, append conflicting claims, then pack everything into token_budget (requested > newer version > conflicts). Cite the returned citations.
| Name | Type | Req | Description |
|---|---|---|---|
| span_ids | array | yes | – |
| token_budget | integer | – | – |
No output schema declared.
No examples provided.
blame ~64
Which version first introduced a passage, and its status (identical/similar/absent) in every version. Give span_id, or text + doc to locate the passage in the latest version.
| Name | Type | Req | Description |
|---|---|---|---|
| doc | – | – | – |
| span_id | – | – | – |
| text | – | – | – |
No output schema declared.
No examples provided.
diff_versions ~105
Compare two versions of a document. Defaults: latest vs the version before it. level: 'section' (added/removed/modified/renumbered sections with changed sentences), 'claim' (requirements whose modality or value changed, added or removed), or 'reference' (citations added/removed).
| Name | Type | Req | Description |
|---|---|---|---|
| doc | string | yes | – |
| from_version | – | – | – |
| level | string | – | – |
| max_tokens | integer | – | – |
| to_version | – | – | – |
No output schema declared.
No examples provided.
find_affected_documents ~61
Documents that reference this document (cite, update or obsolete it). Use it to see what may need review when the document changes; flags sources that still cite an obsolete version.
| Name | Type | Req | Description |
|---|---|---|---|
| doc | string | yes | – |
| max_tokens | integer | – | – |
No output schema declared.
No examples provided.
find_contradictions ~123
Statements from different documents that conflict (different values for the same limit, or opposing MUST/MUST NOT/MAY requirements). Both sides are returned as evidence; nothing is auto-resolved. status: confirmed (rule + LLM judge agree), candidate (rule only) or rejected (judge disagreed, hidden by default). include_history: also include pairs involving obsolete versions.
| Name | Type | Req | Description |
|---|---|---|---|
| doc | – | – | – |
| include_history | boolean | – | – |
| include_rejected | boolean | – | – |
| k | integer | – | – |
| max_tokens | integer | – | – |
No output schema declared.
No examples provided.
get_evidence ~60
Full text of one span with its provenance (byte range, SHA-256, parser version), section path, whether its version is the latest, and a live hash verification result.
| Name | Type | Req | Description |
|---|---|---|---|
| span_id | integer | yes | – |
| verify | boolean | – | – |
No output schema declared.
No examples provided.
get_section_map ~107
Table of contents of a document version: section numbers, titles, levels and page ranges. doc: document id ('employee-handbook'), version id ('employee-handbook@v2') or, for RFCs, 'RFC 5321'. Defaults to the latest version. max_depth: deepest heading level to include (1 = top-level only).
| Name | Type | Req | Description |
|---|---|---|---|
| doc | string | yes | – |
| max_depth | integer | – | – |
| max_tokens | integer | – | – |
| version | – | – | – |
No output schema declared.
No examples provided.
ingest_status ~35
Status of a background add_document job (or the 10 most recent jobs when job_id is omitted).
| Name | Type | Req | Description |
|---|---|---|---|
| job_id | – | – | – |
No output schema declared.
No examples provided.
list_documents ~27
List every document with its versions (oldest→newest) and which version is latest.
Input schema present but exposes no named parameters.
No output schema declared.
No examples provided.
list_versions ~53
All versions of a document (oldest→newest, append-only) with hashes, parser version, origin and declared relations such as 'obsoletes RFC 2821'.
| Name | Type | Req | Description |
|---|---|---|---|
| doc | string | yes | – |
No output schema declared.
No examples provided.
lookup_definition ~58
Find where a term is defined ('resource owner', 'eligible employee', 'mailbox'), with the defining text.
| Name | Type | Req | Description |
|---|---|---|---|
| doc | – | – | – |
| k | integer | – | – |
| max_tokens | integer | – | – |
| term | string | yes | – |
No output schema declared.
No examples provided.
read_section ~100
Read the paragraphs of one section, each with a citation and SHA-256. section: a section number ('4.5.3.1', 'A.1') or part of a title ('Terminology'). include_subsections: also return text of nested subsections.
| Name | Type | Req | Description |
|---|---|---|---|
| doc | string | yes | – |
| include_subsections | boolean | – | – |
| max_tokens | integer | – | – |
| section | string | yes | – |
| version | – | – | – |
No output schema declared.
No examples provided.
resolve_reference ~109
Follow cross-references to their target section, with a preview of the target text. Either span_id (resolve every reference inside that span) or reference text such as 'Section 3.4 of [RFC5322]' together with doc (the document the reference appears in). status: resolved (target stored), external (outside the store) or unresolved (target section missing).
| Name | Type | Req | Description |
|---|---|---|---|
| doc | – | – | – |
| max_tokens | integer | – | – |
| reference | – | – | – |
| span_id | – | – | – |
No output schema declared.
No examples provided.
search ~112
Find passages relevant to a query. mode: 'lexical' (BM25: exact terms, numbers, identifiers), 'semantic' (embeddings: paraphrased questions) or 'hybrid' (both, fused with reciprocal rank fusion). doc: optional filter; a document id searches all its versions, a version id only that version.
| Name | Type | Req | Description |
|---|---|---|---|
| doc | – | – | – |
| k | integer | – | – |
| max_tokens | integer | – | – |
| mode | string | – | – |
| query | string | yes | – |
No output schema declared.
No examples provided.
search_claims ~149
Search normative statements extracted from the documents (sentence-level evidence). topic: words describing what the statement is about ('annual leave carry over', 'text line length'). subject: who it applies to ('employees', 'clients', 'authorization server'). modality: MUST, MUST NOT, SHOULD, SHOULD NOT, MAY (synonyms like SHALL / REQUIRED / must not accepted). claim_type: requirement | exclusion | numeric. Numeric claims include value and unit (e.g. 1000 octet).
| Name | Type | Req | Description |
|---|---|---|---|
| claim_type | – | – | – |
| doc | – | – | – |
| k | integer | – | – |
| max_tokens | integer | – | – |
| modality | – | – | – |
| subject | – | – | – |
| topic | – | – | – |
No output schema declared.
No examples provided.
verify_span ~58
Re-hash a span against the stored source text. verified=false means the evidence was altered and must not be cited. deep=true also re-parses the original file.
| Name | Type | Req | Description |
|---|---|---|---|
| deep | boolean | – | – |
| span_id | integer | yes | – |
No output schema declared.
No examples provided.
How do I install the DocuSense MCP server?
DocuSense runs locally as a PyPI package, launched with uvx docusense-mcp serve-mcp. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
pypi · docusense-mcp
claude mcp add hksharma2004-docusense -- uvx docusense-mcp serve-mcp
What is the DocuSense MCP server?
DocuSense is an MCP server listed in the public MCP registry as io.github.hksharma2004/docusense. Agents read PDFs and docs by section and claim: ~20x fewer tokens, hash-verified citations, diffs. This page covers its PyPI package (docusense-mcp).
Is the DocuSense MCP server safe to use?
DocuSense scores 66 out of 100 on VerifyMCP. 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 DocuSense MCP server expose?
DocuSense exposes 17 tools: list_documents, get_section_map, read_section, search, search_claims, and 12 more. Their descriptions and schemas cost roughly 1,463 tokens of context every time the server is loaded.
Is the DocuSense MCP server still maintained?
DocuSense 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 DocuSense MCP server under?
DocuSense declares the MIT 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 DocuSense MCP server?
VerifyMCP ran 15 checks on DocuSense, covering Known CVEs, Install scripts, Dependency health, Source repository and 11 more. One item could not be completed, a gap on our side rather than a finding about DocuSense: Malware scan. Endpoint Security and Transport & Reachability do not apply to a PyPI package, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.