{"resourceId":"nature-paper2agent-interactive-papers-20260916","versions":[{"version":"external-603645f71399153cb28e8beb6dd935f2d74ad304a0b364ddcd2dc8abf2a13718","resource":{"id":"nature-paper2agent-interactive-papers-20260916","title":"Nature Paper2Agent locks validated paper tools as agents; many repos still fail agentification","organization":"Stanford-affiliated author team (Miao, Davis, Zhang, Pritchard, Zou); Nature","sector":"AI-assisted scientific methods dissemination / university research software","geography":"International (U.S. university authors; transferable to SLED research computing)","publishedAt":"Published 16 September 2026, Nature (DOI 10.1038/s41586-026-11044-y)","publicationDate":"2026-09-16","eventDate":"2026-09-16","sourceName":"Nature","sourceLabel":"Peer-reviewed Nature article; full HTML opened via curl","sourceUrl":"https://www.nature.com/articles/s41586-026-11044-y","evidenceClass":"academic-research","outcomeClass":"mixed","topics":["developers-agents","knowledge-work","operating-model","infrastructure","governance-procurement"],"finding":"Paper2Agent can turn papers into locked, validated MCP agents with strong AlphaGenome query accuracy versus general coding agents, but authors still report a substantial fraction of repositories that cannot be agentified and keep open-ended reasoning human-in-the-loop.","sledRelevance":"Same-week Nature publication on making university research outputs reusable as agents—directly relevant to research computing facilitation, reproducibility services, and AI co-scientist pilots. Distinguishes locked tool reproduction from autonomous discovery.","evidence":"Published 16 Sep 2026. Builds MCP tools/resources/prompts from paper+code; tools validated against reference results/figures then locked. AlphaGenome: 22 tools, ~45 min/$14; 98.7±1.3% tutorial and 100.0±0.0% novel query accuracy (n=5) vs Claude+Repo 82.7%/78.7% and Biomni 37.3%/56.0%; open-ended 82.7% vs 56.7%/72.2%; median runtime cuts ~1.9–3.8×. Scanpy/TISSUE agents and multi-agent psoriasis prioritization shown. Limitation: substantial fraction of repos could not be agentified (incomplete code/docs/environments); open-ended reasoning remains HITL.","architectureImplications":"Interpretation: host paper MCP servers remotely (e.g., Spaces), pin validated tools, and connect campus chat agents via MCP rather than giving models raw repo write access for every query.","governanceImplications":"Interpretation: treat successful agentification and locked-tool validation as reproducibility signals; require human review for novel scientific claims beyond validated tutorials.","securityPrivacyImplications":"Interpretation: remote MCP hosting and API-keyed models still need campus data-class rules; do not feed restricted data into paper agents without authorization.","caveats":"Author benchmarks with human graders on selected queries; agentification failure rate not given as a percentage in inspected text; not a multi-campus production study.","streamIds":["research"],"roles":{"sales":"Interpretation: Engage VPR, research computing, and library/RSE leaders who want papers reused without every lab rebuilding environments. Ask which methods are high-demand for facilitation, what validation evidence they require before endorsing a paper agent, and who owns updates when code drifts. Offer a bounded pilot converting 2–3 well-documented campus methods into locked MCP tools with accuracy checks—not an autonomous discovery promise. Value hypothesis is faster correct reuse of published methods. Unsupported: guaranteed agentification of all repositories or replacement of PI judgment.","engineering":"Interpretation: Fit is research software packaging where methods have code+data+reference outputs. Integrate environment setup, automated tool extraction, validation oracles, remote MCP hosting, and campus IdP-gated chat agents. Prerequisites: reproducible containers and reference figures/results. Constraints: incomplete repos fail; LLM stack choice affects cost/latency. Security: sandbox execution; control API keys; classify data before novel queries. Proof of value: tutorial/novel query accuracy vs Claude+repo baseline and tool lock audits—not publication counts.","delivery":"Interpretation: RSE/library owns packaging; PIs own scientific claims; research computing hosts MCP endpoints. Work: select candidate papers, run Paper2Agent-style conversion, record validation failures, and publish usage guides. Dependencies: GPU/CPU for validation runs and Hugging Face or campus hosting. Train facilitators on when tools are locked vs exploratory. Checkpoints: validation pass before public MCP; HITL for open-ended claims. Acceptance: locked tools match reference outputs; failed agentifications logged. Risks: overselling agentification rates and mistaking query accuracy for new discovery."},"roleDetails":{"sales":{"customerProblem":"Campus researchers struggle to reuse published methods without rebuilding fragile environments, yet leaders fear hallucinated code and overstated AI-scientist claims.","stakeholders":["Research computing leadership","Research software engineers / library digital scholarship","Vice president for research","Method-owning PIs"],"discoveryQuestions":["Which high-demand methods have complete code and reference outputs?","What validation evidence is required before endorsing a paper agent?","Who maintains MCP servers when upstream code changes?","Which data classes may never enter paper agents?"],"valueHypothesis":"Locked, validated paper MCP tools can reduce reuse friction for well-documented methods if failed agentifications are tracked and open-ended claims stay human-reviewed.","potentialEngagement":"Bounded pilot converting 2–3 campus or partner methods into locked MCP tools with accuracy checks against reference outputs.","unsupportedClaims":["All repositories can be agentified","Paper agents replace PI scientific judgment","Query accuracy equals new discovery"]},"engineering":{"fit":"Method packaging into MCP tools with automated validation and remote agent access for genomics/single-cell-style analytical workflows.","architectureIntegration":"Environment agent + extraction + validation loop; deploy MCP remotely; connect campus chat agent via MCP; keep raw repo write paths out of default user flows.","prerequisites":["Reproducible containers","Reference results/figures","Hosting for MCP","IdP-gated access"],"constraints":["Incomplete documentation blocks agentification","Validation cost/time per paper","Model/runtime variance"],"security":"Sandbox tool execution; manage API keys; enforce data classification on novel inputs.","proofOfValue":["Tutorial and novel query accuracy vs baseline agents","Tool lock audit trail","Count of failed agentifications with reasons"]},"delivery":{"work":"Select candidates, convert and validate tools, publish MCP endpoints, and document HITL rules for open-ended queries.","dependencies":["PI method owners","Campus hosting","Compute for validation","Research compliance"],"ownership":"RSE/library maintains MCP packaging; PIs own scientific interpretation; research computing operates hosting.","skillsAdoption":"Train facilitators to distinguish locked validated tools from exploratory agent coding on raw repos.","governanceCheckpoints":["Validation pass before public release","Periodic re-validation on upstream changes","HITL required for novel scientific claims"],"acceptanceCriteria":["Locked tools match reference outputs within tolerance","Failed conversions logged with reasons","Access controls documented"],"risks":["Overstated agentification success","Stale MCP after upstream fixes","Sensitive data entered into paper agents"]}},"retrievedAt":"2026-09-18T03:07:30Z","enrichedAt":"2026-09-18T03:10:00Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Interpretation: natural-language paper agents can widen method access for non-programmers if training covers when locked tools apply and when human expertise is required.","procurementImplications":"Interpretation: evaluate paper-agent platforms on validation locking, traceability to source code, and failed-conversion reporting—not demos alone.","operatingModelImplications":"Interpretation: separate packaging/hosting ownership from scientific acceptance; treat agentification difficulty as a reproducibility signal.","sourceVerification":{"openedUrl":"https://www.nature.com/articles/s41586-026-11044-y","referenceExcerpt":"Each tool used by a paper agent is validated against the reference codebase’s reported results","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}