Review an authorized academic manuscript through literature positioning, novelty, logic, methods and statistics, factual consistency, citation and hallucination checks, presentation quality, and an integrated evidence-traceable review.
yangheng95/opencorvus
- Latest recorded source activity
- SkillsMP catalog refreshed
- skills collected
- 145
- GitHub stars
- 312
- GitHub forks
- 40
Install with an AI assistant
Copy this prompt into the AI assistant you're using.
Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/yangheng95/opencorvus.Loading the repository README…
Skills in this repository
Showing 40 of 145 collected skills.
Execute the binding commercial legal evidence chain with one typed Artifact per owner, two parallel analyses, independent review, and canonical report delivery.
Execute the binding Data Analysis & Business Insights evidence chain with exact Artifact discovery, complete reads, explicit source selection, parallel analysis, audit, and canonical publication.
Execute the binding multi-perspective research chain through exact Task Artifact discovery, complete reads, explicit selections, independent review, and canonical publication.
Generate and execute the smallest Task-local Agent team and workflow description directly through existing OpenCorvus Sessions.
Execute the binding Equity Research evidence chain through exact Task Artifact discovery, complete reads, explicit selections, and one typed output per owner.
Turn a product brief into an original evidence-backed desktop interface through source research, explicit design direction, implementation, and rendered review.
Execute the binding Human Resources & Organization Operations evidence chain with exact Artifact discovery, complete reads, explicit source selection, parallel analysis, audit, and canonical publication.
Execute the suggested Marketing & Growth Strategy workflow with typed Artifact handoffs and real parallel branches.
Execute the binding omnichannel delivery-pack chain with two parallel evidence frontiers, one joined plan, one-predecessor readiness review, and Build-owned publication.
Execute the binding Sales Strategy & Customer Research evidence chain with exact Artifact discovery, complete reads, explicit source selection, parallel analysis, audit, and canonical publication.
Execute the suggested SEO & Generative Engine Optimization workflow with typed Artifact handoffs and real parallel branches.
Author one portable, project-owned and traceable OpenCorvus Expert Squad through the Generate Expert Squads workflow and canonical SDK writer.
Execute the binding tax compliance evidence chain with parallel accounting and tax analysis, independent review, and canonical reporting.
Execute the binding Viral Content Artifact chain through exact discovery, complete reads, semantic selections, independent review, and Build-owned publication.
Execute Advanced software delivery through requirements-to-architecture traceability, safe parallel evidence frontiers, exact implementation ownership, and independent acceptance.
Build trustworthy quantitative evidence and turn accepted findings into one integrated, browser-reviewed research report. Use in Research Studio for dataset analysis, calculation-heavy research, dashboards, standalone HTML reports, or any delivery whose…
Deliver one authorized outcome with an accountable owner and independent evidence comparison.
Research and operate rendered web pages through OpenCorvus Browser tools, including authenticated workflows and visual inspection.
Operate native desktop apps through OpenCorvus Computer tools when the task requires their visible interface.
Execute an explicit Expert Squad incumbent-challenger campaign with immutable package revisions, exact run evidence, frozen scorers, independent integrity review, and a non-executing recommendation.
Coordinate an operator-defined Mission through native Mission state and fixed-profile Tasks without imposing a domain-specific workflow.
Run reproducible OpenCorvus harness experiments against AutomationBench, preserve paper-grade evidence for every attempt, invalidate bug-affected runs, and maintain a self-owned leaderboard. Use for AutomationBench runs, reruns, score cataloging,…
Run reproducible OpenCorvus Mission/Base experiments against Tencent WorkBuddy Bench, require the runtime Code Skill, preserve official Harbor grading and paper evidence, and exclude bug-affected attempts. Use for WorkBuddy Bench setup, chain proofs, runs,…
Install, configure, verify, operate, and troubleshoot OpenCorvus from an assistant such as Hermes Agent, OpenClaw, Codex, or another Agent Skills-compatible host. Use when the user asks to install OpenCorvus, connect a model provider, run OpenCorvus locally,…
Agent-led heterogeneous Squad discovery, evidence review, blocker repair, lossless Skill-directory and portable remote MCP preservation, and canonical import into OpenCorvus.
Author one portable, project-owned and traceable OpenCorvus Expert Squad through the Generate Expert Squads workflow and canonical SDK writer.
Run bounded read-only consultation, investigation, option analysis, planning, and clarifying-question work with variable parallel Planner and Investigator sessions. Use when the requested result is knowledge or advice rather than implementation or an external…
Produce a new evidence-backed editable PPTX through one bounded plan, parallel source validation and delivery construction, and the qualified Work Artifact lifecycle. Do not use for DOCX, XLSX, PDF, existing-presentation editing, or multi-format packs until…
Create, inspect, validate, review, and deliver enabled structured Work Artifacts through the single OpenCorvus Work Artifact Harness. Use for presentation and other document, PDF, spreadsheet, data, image, archive, OCR, or media requests only when the mounted…
Plan and execute bounded real-browser research and UI acceptance with current page evidence, safe interaction boundaries, and independent parallel observation and criterion-checking workers. Use when rendered state, navigation, or visible interaction behavior…
Build source-bound campaign measurement, delivery-quality, attribution, experiment, and brand-suitability evidence for qualified human review without live advertising-platform mutation or compliance conclusions.
Builds source-bound clinical estimand, SAP, population, SDTM/ADaM derivation, model, missing-data, multiplicity, sensitivity, interim, blinding and DMC evidence packs. Use for qualified statistical review, never unblinding, threshold selection, result…
Prepare source-bound facility fire-protection basis, passive and active system, supplied model, inspection, testing, and impairment evidence for qualified review. Use for bounded building or facility fire/life-safety assurance where compliance, design,…
Evidence-first internal-audit method for charter-bound audit universes, risk prioritization, control design and walkthroughs, population and sample testing, operating effectiveness, exceptions, findings, root cause, remediation and closure verification. Use…
Builds source-bound medical-device installed-base, complaint, adverse-event, vigilance, trend, PMCF, RWE, field-action, CAPA and risk-file evidence packs. Use for postmarket review preparation, never causality, reportability, recall, compliance, filing or…
Source-bound M&A due-diligence method for authorized deal perimeter, VDR completeness, financial-quality and working-capital reconciliation, commercial/customer/market/operations evidence, and legal/regulatory/technology/people issue extraction. Use when a…
Prepare source-bound ocean platform, instrument, calibration, profile, time-series, QARTOD, multidimensional-format, coordinate, provenance, and cross-platform validation evidence. Use for bounded oceanographic data assurance without sensor control, flag…
Prepare source-bound electric-grid topology, protection-zone, relay-setting, fault-study, disturbance, misoperation, outage, and reliability-definition evidence for qualified review. Use for bounded protection and reliability assurance without relay, SCADA,…
Builds source-bound radiotherapy equipment, commissioning, reference dosimetry, machine-quality, TPS, patient-specific QA, change, incident and independent-audit evidence packs. Use for qualified medical-physics review preparation, never treatment planning,…