Back to directory

dsh-files

Curated pickMaintenance: Active

taxueseek/dsh-files

DSH 双面文件插件(会话隔离上传 + read_document 文档读取,内容嗅探)

View on GitHub
$ curl -fsSL https://raw.githubusercontent.com/taxueseek/dsh-files/main/install.sh | sh

40

stars

11

forks

TypeScript

Language

MIT

License

2026-08-14

Created

2026-10-01

Last push

Native DeepSeek Harness plugin adding the `read_document` tool for PDF / DOCX / XLSX extraction that the built-in read tool cannot handle. MIT-licensed, harness 0.1.3+.

DSH integration

Native runtime

Author-claimed

Safety audit

Unaudited

Last verified

2026-09-06

License

MIT

01What can it help you accomplish?

  • Read binary documents (PDF / DOCX / XLSX) that the built-in read tool rejects

    Structured text extraction plus encoding fallback, paging, and sheet-level XLSX access the model can consume

    DeepSeek Harness users who need to ingest binary documents (PDF / DOCX / XLSX) into the model

  • Read a specific Excel worksheet from an uploaded workbook

    `list_sheets` names the sheets and the `sheet` parameter reads one sheet in full (no row cap); out-of-range errors list available sheets

    DeepSeek Harness users working with spreadsheet data who need sheet-level, not whole-file, access

02How to install into DeepSeek Harness

Prerequisites

  • Requires harness ≥ 0.1.3-alpha.1.

Installation steps

  1. 01

    $ curl -fsSL https://raw.githubusercontent.com/taxueseek/dsh-files/main/install.sh | sh

  2. 02

    # restart dsh web

  3. 03

    $ dsh plugin --profile web add git+https://github.com/taxueseek/dsh-files.git

  4. 04

    # restart dsh web

Verify the integration

Not specified by the author

03DSH integration and capability boundaries

DSH integrationNative runtime

Native dsh plugin: adds the `read_document` tool that fills the PDF / DOCX / XLSX extraction gap left by harness 0.1.3's built-in read tool, reading through `ctx.fs` and inheriting the session sandbox

  • Content sniffing

    bytes of an uploaded document→format decided from bytes (PDF header / ZIP members / UTF-8 / UTF-16 BOM / GB18030); disguised files rejected

  • Encoding chain

    text bytes of unknown encoding→UTF-16 BOM → UTF-8 (fatal) → GB18030 (fatal) → UTF-16 without BOM; GBK and BOM-less UTF-16 both read

  • Paged reads

    a document handle + line offset/limit→line numbers + offset/limit; per-call character budget differs by format, overflow truncates with a remaining-lines marker

  • XLSX sheet-level reads

    an uploaded .xlsx workbook→`list_sheets` names sheets; `sheet` parameter reads one sheet in full (no row cap); out-of-range errors list available sheets

  • Cooperative cancellation

    an in-flight parse→parsing aborts immediately on user cancel / session close

04Who is it for? When not to use it?

Good for

  • DeepSeek Harness users who need to ingest binary documents (PDF / DOCX / XLSX) into the model
  • DeepSeek Harness users working with spreadsheet data who need sheet-level, not whole-file, access

Not for

  • The npm package named dsh-files is an unrelated third-party placeholder — install only via the official install script or the `dsh plugin add` git command, never via `npm install dsh-files`.

05Compatibility, maintenance and safety notes

  • The npm package named dsh-files is an unrelated third-party placeholder — install only via the official install script or the `dsh plugin add` git command, never via `npm install dsh-files`.
  • A PDF with no text layer returns an explicit notice, not extracted text — dsh-files does not perform OCR on scanned documents.
  • Requires harness ≥ 0.1.3-alpha.1; older harness versions are not supported.
Not specified by the authorNot specified by the authorv0.5.1

readme_verified

06Frequently asked questions

What does dsh-files do that the built-in read tool can't?

The built-in read tool rejects binary content with `FS_NOT_TEXT`. dsh-files adds the `read_document` tool for structured text extraction of PDF / DOCX / XLSX, plus encoding fallback, paging, and sheet-level XLSX access.

How do I install dsh-files?

Run `curl -fsSL https://raw.githubusercontent.com/taxueseek/dsh-files/main/install.sh | sh`, or `dsh plugin --profile web add git+https://github.com/taxueseek/dsh-files.git`, then restart dsh web. Requires harness ≥ 0.1.3-alpha.1.

Is the npm package named dsh-files the same plugin?

No. The npm package named `dsh-files` is an unrelated third-party placeholder — install only via the script or the git command above.

What happens with scanned PDFs?

A PDF with no text layer returns an explicit notice, not an empty string. dsh-files does not perform OCR on scanned documents.

Which harness version is required?

Requires harness ≥ 0.1.3-alpha.1.

08Data and sources

  • Author-claimedgithub.coma814c7b89b08…

    A DeepSeek Harness plugin that does exactly one thing: the **`read_document` tool** — structured text extraction for bin…

  • Author-claimedgithub.coma814c7b89b08…

    Reads go through `ctx.fs`, inheriting the session sandbox, same rights as the built-in read tool

This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-09-06. Found an error? Submit a correction.

🏆

Best DeepSeek Harness Plugins

Twelve plugins worth installing first — picked from the whole catalog, across every category.

DSH Plugins is an independent community directory of DeepSeek Harness plugins. Not affiliated with or endorsed by DeepSeek. Third-party plugins are not security-audited — review the source before installing.

New DeepSeek Harness plugins, weekly. No spam.