dsh-research-suite

A DeepSeek Harness (DSH) plugin for the full clinical-research workflow — manuscript writing, language polishing, figures, statistics, research ideation, peer review, slides, literature search and journal formatting. Install it and it takes over research tasks automatically: no UI, no manual skill picking.

中文完整说明见 README.zh-CN.md（内容更全，含配置项、目录结构、占位符清单）。

• Package: dsh-research-suite · Version: 2.0.0 · License: MIT
• Requires Node ^22.19.0 || >=24.0.0, DSH >=0.2.0-0 <0.3.0
• Self-contained: all 719 original skill documents ship inside the package; nothing is read from any external skill directory at runtime.

Why capability cards instead of a skill router

The hard part of research work is not a lack of skills — it is that three to eight skills can each do the same task well in different ways. A router hands the model one "primary skill" name; the model then reads that skill's manual and improvises, losing the other skills' strengths along the way.

Instead, each capability card takes every competing skill in its group, records best_at (what it is best at), limits (what it explicitly cannot do) and take (the concrete rules, thresholds and templates lifted from it), then fuses them into one executable protocol. Every protocol step says which skill it comes from and why. The acceptance rule is strict: after reading one capability card you can finish the job without opening any original skill document.

The 719 original documents remain bundled as a deep-reading layer — for checking details, not for improvising.

How auto-takeover works

No UI, no buttons. Three pieces:

1. A system-prompt section — lib/prompt.js builds a policy named research-suite/policy at order: 650 ("after team policy, before tool docs"); lib/index.js registers it through systemPrompt.section({ name, order, text: () => buildPolicy(core) }). The policy tells the model: recognise a research task → call research_playbook first → follow the capability protocol → deep-read if needed → run research_gate before delivering.
2. 10 skills (skills/): one entry skill plus nine domain SOPs. Their content is inlined at registration time — passing a lazy load() hook fails DSH's runtime validation and silently drops every skill (the exact failure mode of v1).
3. 7 read-only tools for capability lookup, searching the 719-document catalog, deep reading, standards, and mechanical self-checks.

When research_playbook finds no match, the policy falls back to research_catalog + research_kb_read rather than refusing the task.

Domains

Domain	Skills	Capabilities	Covers
writing	51	11	paper types and sections (original / review / case report / grant)
polish	24	5	register, de-AI-flavour, reduction, proofreading, Word reports
figures	76	9	data plots, statistical charts, mechanism diagrams, layout
stats	180	16	descriptive, comparison, regression, survival, prediction, PSM, MR, Meta, missing data, sample size
ideation	33	7	topic selection, novelty search, hypotheses, grants, ethics
review	39	8	self-review, reviewer comments, revision, rebuttal, response letters
slides	44	7	conference, lecture, defence, lab meeting, poster
literature	152	11	search strings, screening, close reading, evidence tables, recall/precision
format	32	7	journal formatting, references, figure/table specs, submission package
Total	631	81	81 capability cards across the nine domains (177 skills formally referenced, 542 catalogued for on-demand reading)

631 skills across the nine domains plus 88 general skills = 719 bundled skill documents (~9.9 MB). Nine reference standards (Chinese Medical Association format, GB/T citation, statistical reporting, figure specs, grant rules, etc.) live in references/.

Tools

Tool	Purpose
research_playbook	The only entry point. Pass the user's own words; returns the 1–3 best-matching capability cards with protocol, gates, deliverables and suggested deep reads
research_capability	Expand a full capability card (member comparison, protocol, decisions, gates, outputs, kb) by id or by domain
research_catalog	Search the 719-document catalog by name/description keywords, optionally filtered by domain
research_kb_read	Read one skill document or INDEX.md/FUSION.md, with paging and path-escape rejection
research_standard	List or read the nine reference standards in references/
research_gate	Gate and mechanical self-check: section = all / data / selfcheck / mutex / forbidden / hard / handoffs; text checks fullwidth punctuation, superscript citation numbers, AI-cliché phrases and absolute wording
research_suite_status	Load self-check: version, per-domain capability counts, catalog completeness, references, package root

Install

# via plugin_manager (GUI)
plugin_manager(action="install_bundle", target="<absolute package path or dsh-research-suite@2.0.0>")

# via CLI (Windows)
dsh.cmd plugin --profile desktop add <absolute package path or package name>
# via CLI (POSIX)
dsh plugin --profile desktop add <absolute package path or package name>

Fully quit and restart DSH Desktop afterwards (closing the window is not enough — the loader reads the bundle patch layer at startup). Verify with research_suite_status.

Configuration

Key	Default	Meaning
autoTakeover	true	Inject the auto-takeover prompt section; false degrades the plugin to "7 tools + 10 skills" without automatic routing
sectionOrder	650	Order of the prompt section (after team policy, before tool docs)
verbose	false	Log warnings when skill/tool/prompt registration fails

Documentation

• docs/CAPABILITY-SPEC.md — capability-card content contract
• docs/ARCHITECTURE.md — modules, data flow, how auto-takeover is implemented
• docs/PUBLISH.md — npm release, DSH install paths, upgrade/rollback, troubleshooting table
• CHANGELOG.md — what changed in 2.0.0

Self-checks

npm test           # scripts/selftest.mjs — fake-ctx load smoke test
npm run verify     # scripts/verify-package.mjs — static release checks
npm run verify:caps# scripts/verify-capabilities.mjs all
npm run pack:check # npm pack --dry-run (~11-12 MB unpacked / ~4 MB tarball; knowledge/ included by design)

Measured on the 2.0.0 release candidate: 773 files / 11.24 MB on disk (scripts/verify-package.mjs count, includes .github/); npm pack ships 771 files, 11.8 MB unpacked, 4.1 MB tarball.

License

MIT — see LICENSE.

Scope note: MIT covers the plugin itself (lib/, data/capabilities/, skills/, references/, scripts/, docs/ and the index files). The 719 skill texts under knowledge/skills/ are third-party material vendored verbatim from the author's local skill collection; their licenses belong to their original authors and are not covered by MIT. Read docs/ATTRIBUTION.md before redistributing.