Academic Harness

A research workbench for DeepSeek Harness — literature and web discovery, verbatim paper memory, a writing oracle for LaTeX, systematic-review, meta-analysis, and reproducible-analysis ledgers, 774 bundled skills, and local self-improvement, all inside the DSH web UI.

Academic Harness is a DSH profile: it installs as a plugin, adds a preset with a research persona and academic tools, and keeps the general-purpose tools available on every preset.

Why Academic Harness

Principle	What it means in practice
Evidence, not citation-shaped guesses	PaperMemory stores papers as content-addressed Markdown and anchors claims to exact quote spans. quote_check and trace verify a quotation before it becomes a citation; an empty cite means do not invent one. Read more
A writing oracle with a gate	The vendored ah_* engine inventories every number, citation, unit, and claim in a LaTeX tree. A turn closes only when it introduced no open blocker or major. Read more
Auditable research ledgers	Systematic reviews and hypothesis tournaments live in append-only logs under the workspace, so a resumed session sees the same protocol, screening decisions, and ratings. Reviews · Tournaments
Discovery across the real literature	OpenAlex, Semantic Scholar, PubMed/Europe PMC, arXiv, Crossref, DOAJ, publisher Crossref deposits, bioRxiv/medRxiv, INSPIRE, and keyed ADS/Scopus/Web of Science through one literature_search entry point, plus journal_search, paper_get, paper_context, paper_resolve, openalex_citations, database_catalog, domain_search, Consensus, Parallel, Firecrawl, Scite, CORE, Elicit, and FutureHouse. Read more
A lab notebook that survives context loss	A JSONL notebook shared across the delegation tree, a schema-validated scientific_result card, and deletion-only reply folding at five levels. Read more
One question, one bounded investigation	deep_research reads the request, sweeps the literature and a web lane, finds where the sources disagree, drafts, lets four critics attack the draft, and refuses to ship a report that cites a source it cannot support. Five syndicated copies of one release count as one voice. `effort=low
Local-first by design	Workspace search (zvec-grep), PaperMemory, TextFold, ledgers, and the RSI store stay on disk. No key is required to boot. Read more

Install

Requires DeepSeek Harness 0.2.1–0.2.x (>=0.2.1-rc.5 <0.3.0, the range in package.json) and Linux (the only platform tested so far; see Requirements).

npm install -g @deepseek-ai/dsh

Then either:

A — from the DSH Web UI (recommended): open Plugins → Add plugin, paste github:raktim-mondol/academic-harness, click Install, and start a new chat.

B — a dedicated CLI profile:

dsh plugin --profile academic-harness add github:raktim-mondol/academic-harness
dsh --profile academic-harness

Confirm the preset selector at the top of the message box reads Academic Harness. The academic tools are preset-only; a session left on Standard/PTC will fail with tools[name] is not a function.

Full instructions, troubleshooting, and the "one DSH surface at a time" rule are in Install and run.

What is inside

Surface	Highlights	Guide
Search and discovery	literature_search router, journal_search (OpenAlex/DOAJ/Crossref metrics plus optional CiteScore/JIF), paper_get (merged metadata + BibTeX), paper_context (focused sections/snippets), database_catalog, domain_search (bioRxiv/medRxiv, INSPIRE, ACL, ADS, Scopus, Web of Science, Wikidata), OpenAlex, Semantic Scholar, PubMed, arXiv, Crossref, DOAJ, publisher Crossref deposits, Consensus, Parallel, Firecrawl, Scite, Tavily, Exa, Perplexity, Jina, CORE, Elicit, FutureHouse	Search
Papers and evidence	paper_get, paper_download, paper_context, paper_fold, pdf_to_markdown, deep_review (multi-pass manuscript critique with anchored comments and reference verification, powered by open-deepreview), PaperMemory's 14 tools, reference_library (the native reference manager: BibTeX, BibLaTeX, RIS, and CSL-JSON in and out; imports Scopus and Web of Science RIS exports), integrity_check, Paper2Agent	Papers
Writing and citation	ah_* writing oracle and gate, citation (including DOI/arXiv/INSPIRE/title → BibTeX, batch audit, and repair that writes a new .bib), manuscript_citations, reporting_check (CONSORT/STROBE/PRISMA/ARRIVE/TRIPOD+AI/CARE/STARD), LaTeX, Supervisor-Skills paper lifecycle, scientific-figure-making, diagram-design, image_generate, mermaid_render (Pele)	Writing
Submission and communication	rebuttal review-point coverage and evidence-linked responses; submission_pack journal checklist and hashed bundle (never submits); journal_search match advisory venue fit with due-diligence signals; manuscript_diff revision artifacts and response-letter skeleton; numbers_audit numerical-reporting candidates; presentation_build posters and slides with a vision-QA checklist; authorship CRediT/ICMJE ledger	Writing
Reviews, synthesis, and hypotheses	review PRISMA-style ledger with dual-reviewer screening, structured extraction, risk-of-bias appraisal (RoB 2, ROBINS-I, QUADAS-2, Newcastle-Ottawa, AMSTAR 2, PROBAST, QUIPS, SYRCLE, CASP checklists), GRADE certainty and summary-of-findings tables, κ agreement, PRISMA flow and RoB exports; meta_analysis pairwise pooling, heterogeneity, subgroup/meta-regression, publication-bias diagnostics, and forest/funnel/SROC plots; coscientist Glicko-2 tournament, published-results feedback, cross-session priors, append-only limitation records with coverage, experiment traces fed back into evolution, evidence-bound champion selection, and grounded novelty via coscientist_verify	Research
Deep research	deep_research: a self-driving investigation on an append-only run ledger — frozen query, effort levels (low/medium/high) inferred from the request, scholarly sweep plus a web lane, contradiction graph, loci and depth notes, three drafts, four adversarial critics, capped-hunk repair, cite check against PaperMemory, independence clustering, untrusted-source fencing, and a ship gate that reports not_checked honestly	Research
Reproducible analysis and data	analysis_run frozen plans, hashed manifests, declared numbers, re-execution checks, and frozen-rule compliance; experiment_log staged subset/full comparisons with named reproduced or reported references, protocol drift, verdicts and Pareto ranking; ablation linked removal arms and attribution; dataset_get, data_profile, and repro_package with compliance exclusions and confirm-gated Zenodo drafts	Reproducible analysis
Provenance audit	provenance_audit read-only cross-ledger index: joins review, reporting, deep-review (claims, citation verdicts, validity flags), notebook, analysis (expectations, re-executions, compliance), experiment, ablation, dataset, meta-analysis, and AH claim sidecars by stored ids and hashes; coverage, unsupported (no_link/not_checked/unverified/stale/not_found/spec_drift/no_attribution/rerun_mismatch/verdict_missing), stale hashes, impact, the four-column arms readiness roll-up, the report-only rollup loop record with a reproduced-baseline gate and honest cost/time, and Markdown/CSV/JSON/GraphML/Mermaid exports	Provenance audit
Scientific data and toolkits	science_source (UniProt, RCSB PDB, AlphaFold DB, PubChem) with provenance on every record; local chem_toolkit (RDKit), seq_toolkit, and structure_toolkit that log hashed inputs	Scientific data
Discovery and reference checks	literature_watch saved searches with unseen-only digests, paper_recommend, citation_network (frozen OpenAlex sample, centrality and clusters), paper_artifacts link/accession extraction, citation batch audit	Discovery
Identity, funding, and audits	author_lookup (ORCID/ROR), openreview, grant_search, patent_search, stat_audit, figure_audit, repo_audit (including probe and method→code align), pdf_references (GROBID); advisory, never verdicts	Identity and audit tools
Study planning	power_analysis: sample size, power, effect sizes, multiplicity correction, simulation	Research
Study conduct, ethics, and grants	preregister hashed write-once plans; causal_dag identification diagnostics and E-values; design seeded randomization/DOE; ethics_protocol and deidentify for ethics and PHI/PII review; dmp funder-shaped data-management plans; grant_budget multi-year budgets with arithmetic checks. None submits or certifies compliance	Research
Clinical and policy records	clinicaltrials, openfda, dataset_search, journal_policy, funder_policy	Research
Workspace and notebook	zvec-grep semantic search, lab notebook, scientific_result, fold_text, research_template, sci_inspect	Workspace
Compute and delegation	runpod_run/modal_run/vastai_run, subagent with optional subagent_pro/subagent_vision routes, specialist briefs, agent-browser	Compute · Agents
Skills	774 bundled skills: 140 domain, 16 methodology, 503 expert profiles, 23 drug-discovery, 12 Supervisor-Skills research-lifecycle skills, plus writing, figures, agents, and compute; pele renders Mermaid without a browser	Skills
Self-improvement	Local RSI: capture, diagnose, admit, inject, compare against control, promote or quarantine	RSI
Configuration	Settings → Academic Harness API keys; every credential is optional at install	Configuration

Search tools

Tool	What it does
literature_search	One-shot multi-source router with RRF fusion and cross-provider dedupe
journal_search	Venue discovery and metrics across OpenAlex, DOAJ, and Crossref; optional CiteScore/SJR/SNIP and JCR JIF/JCI
paper_get	Merge one paper across OpenAlex, S2, Crossref, Europe PMC, and arXiv; OA URL and canonical BibTeX
paper_context	Focus-ranked sections/snippets from S2, arXiv HTML, PMC BioC/PubTator3, optional publisher TDM, and OA PDF
paper_resolve	Convert DOI/PMID/PMCID/arXiv/OpenAlex/S2 and source-specific INSPIRE/ACL/bibcode/Scopus/WoS/Wikidata identifiers
openalex_search	OpenAlex works, authors, sources, and topics with filters, facets, and source metrics
openalex_citations	Forward, backward, or both citation trails with an OpenCitations fallback
s2_search	Semantic Scholar relevance, bulk, and paper-scoped snippet search
pubmed_search	Europe PMC by default; NCBI E-utilities on request; preprint filter
arxiv_search	Fielded arXiv queries, serialized to one request every 3 seconds
crossref_lookup	Canonical DOI metadata; polite pool via CROSSREF_MAILTO
domain_search	Source-specific bioRxiv/ACL/INSPIRE/ADS/Scopus/WoS/Wikidata actions
database_catalog	Source/key/capability catalog; reports key fields only as configured/not set

Free sources need no key. ADS, Scopus, Web of Science, Springer, IEEE, and publisher TDM tiers use their own keys and are implemented but intentionally not live-tested here. Web and AI-native discovery remains available through consensus_search, parallel_search/parallel_extract, brave_search, tavily_search, exa_search, perplexity_search, jina_read, core_search, elicit_search, and futurehouse_agent, plus the Parallel, Firecrawl, and Scite MCP connectors. See Search for routing and Configuration for the key fields.

The complete tool list, with scope and links, is the tool index. The architecture guide explains which tools are available on every preset and which are preset-only.

How a request works

Discovery finds candidate papers, paper_resolve collapses their identifiers, paper_get produces one merged metadata/BibTeX record, and integrity_check reports retraction status. A downloaded PDF is stored as content-addressed Markdown in PaperMemory, and a quotation must pass quote_check or trace before the agent treats it as evidence. The notebook records the work; scientific_result records the final finding.

Evidence-led research and revision

The ScientistTwo-inspired roadmap has implemented phases 0–10; maintenance continues. Its components support a human-directed, auditable workflow, not an unattended loop:

1. Ground ideas: coscientist records source-quoted limitations, prior-art checks, idea coverage, experiment traces, and an evidence-bound selection.
2. Test and attribute: experiment_log distinguishes reproduced from reported baselines and records subset/full stages, protocol drift and verdicts. analysis_run records reproducible runs and score checks; ablation links removal arms to evidence.
3. Review and revise: deep_review preserves rounds, compares changes and records evaluator provenance. rebuttal tracks reviewer points, linked evidence, answers and declines; a fresh review can check what changed.
4. Audit the record: provenance_audit joins existing ledgers, reports the four integrity columns and a read-only rollup, including missing evidence and unrecorded cost. Violations of frozen task rules are excluded from selection and reproducibility packages.

Readings and verdict records are advisory, not acceptance decisions. The roll-up does not execute experiments or promote an incumbent; not_checked is not a pass. The existing ah_check writing gate remains separate. See Research, Analysis, Writing, and Provenance.

Deep review

deep_review runs a full Deep-Review-style critique of one manuscript with the vendored open-deepreview engine. The paper is segmented and every segment is reviewed with the whole manuscript in context; comments are anchored to a page or a quotation, and a scholar-peer pass adds strengths and weaknesses, a claim-evidence map, a nine-item reproducibility checklist and questions for the authors. estimate prices a run before it starts, review runs on the session model (no separate key or endpoint), show reads a stored review back, and status marks it stale after the manuscript changes. Every pass is kept as a numbered round, so reviewing again after a revision does not erase the first, and compare diffs two rounds — findings closed, opened, or moved in severity, plus claim, checklist, and citation transitions. validity reads a stored round against a deterministic count of the manuscript itself and answers four advisory questions: whether a claimed gain has a removal arm behind it rather than only generic training controls, whether a stated before-and-after change sits in the evaluation harness rather than the model, how wide the results grid actually is, and whether a quality or feasibility counter-metric is reported beside a speed or memory gain. Each answers clear, flagged, or not_checked — a flag names an absence in the text, not a flaw in the work, and not_checked is not a pass — with at most one critic model call on the session route and none with critic: false. score adds four advisory dimension readings — soundness, originality, clarity and significance — on a declared scale, each citing stored comment ids; it is one bounded model call, never an acceptance decision or a gate. Rounds record whether the reviewer was held_out or in_distribution, and Phase 0 can correlate stored scores with real decisions. Artifacts stay in .dsh/deep-review/<key>/, and provenance_audit indexes the claims stored there together with extra_checks.json (a flagged validity check appears as an unverified claim; clear is not a claim). Read more

Deep research

deep_research runs the investigation a literature question actually asks for, and keeps the state a chat cannot. The request is frozen verbatim and re-read by every step, so a long run cannot quietly drift onto a different question. It sweeps the literature through the existing literature_search router and a web lane (Parallel, Exa, Tavily, Brave) for guidance documents, registers, and reporting; finds where the sources contradict each other; picks the loci where reading deeply would change the answer; drafts from three angles; then has four critics attack the draft and repairs it with capped hunks — after synthesis the report cannot be regenerated, only patched.

Two things it counts that a plain agent gets wrong. Voice, not copies: an independence clustering pass groups DOI, exact and near titles, and identical bodies across hosts, so five reprints of one press release argue with the weight of one source (distinct DOIs are never merged). Verified, not plausible: every citation is checked against stored text in PaperMemory, a quote nothing read is reported not_checked rather than passed, and an unacknowledged retraction blocks the ship. Fetched pages arrive inside an <untrusted-source> fence, so a page that tries to instruct the agent is read as content.

Scale is one knob — effort=low|medium|high — and it is inferred from the request when not given ("quickly" is low, "comprehensive systematic review" is high). A request that points both ways returns needs_input with a question and creates nothing: no run, no model call. Runs are resumable and budget-capped, so a crash continues at the step that died instead of restarting. The pipeline shape — step-per-phase, the frozen canonical query, patch-only repair, the ship gate, independence clustering, and untrusted-source fencing — is adapted from hyperresearch (MIT); no upstream code is vendored, and NOTICE records the attribution. Read more

Documentation

Start here	Go deeper	Reference
Install and run	Search	Tool index
Architecture	Papers and evidence	Configuration
Documentation home	Writing	Changelog
	Research	Development
	Workspace	Engine notes
	Compute · Agents · Skills · RSI
	Analysis · Provenance · Scientific data · Discovery · Identity and audit tools

Requirements and scope

• Tested on Linux only. The plugin, its runtime installs (PaperMemory, zvec-grep, TextFold), and the end-to-end browser tests have been run on Linux (Python 3.13, Node 24). macOS and Windows have not been tested and may not work; Python is needed for PaperMemory, power_analysis, the sci_inspect helpers, and the managed Modal SDK.
• DSH 0.2.1–0.2.x. The plugin's bundle patches the web-app layer, so the profile must include @deepseek-ai/dsh-web-app; the standard web profile does.
• A preset check matters. The research persona, search connectors, PaperMemory, zvec-grep, paper tools, citation/review/tournament tools, and the subagent_pro/subagent_vision routes are scoped to the Academic Harness preset. Both routes inherit the session model unless you configure a provider/model. Only the general-purpose tools (notebook, fold_text, scientific_result, sci_inspect, images, LaTeX, PDF, Mermaid, compute, templates) are on every preset.
• Keys are optional. No credential is needed to install or start; each connector stays disabled until its key is present. Some runtimes install in the background on first use — see Architecture.
• One DSH surface at a time. dsh-web, dsh-tui, dsh-martty, and academic-harness share $DSH_HOME, port 3080, and the zvec-grep daemon. Stop the previous one before starting another; see Install.

Development

npm install       # once, for the Settings page build
npm run build     # after any client/ change
npm run docs:diagrams   # regenerate the SVG figures from docs/diagrams/*.mmd

Server-side code is plain ESM JS with no build step. Tests are plain Node scripts; see Development for the full suite and the paper_fold live test.

License

MIT — see LICENSE. Scientific skills are vendored or adapted from several open-source catalogues; NOTICE records exactly what was copied, renamed, or adapted from each. The Supervisor-Skills bundle is mixed-license: six skills are CC BY 4.0, five are CC BY-NC-SA 4.0 (non-commercial, share-alike), and drawio-reconstruction is MIT; those terms apply to those skill files, not to the plugin as a whole.