RW Review Methods

Startup

1. Read

   references/standalone.md

   ,

   references/method.md

   , and

   references/standards.md

   .
2. Read the materials provided by the user in this round; if no materials are provided, only complete the parts allowed by the current evidence.
3. Retrieve

   references/atoms.jsonl

   when rule judgment is required; read

   references/cases.md

   when encountering similar tasks.
4. Read

   references/domain-guide.md

   when selecting methods or tools, and use

   assets/worksheet.md

   to organize deliverables.
5. Read

   references/acceptance.md

   .

   references/behavior-tests.json

   is only used for testing and not as factual information for user tasks.
6. When current literature, APIs, report specifications, and journal requirements may change, verify and record the date by opening the official links in

   references/source-map.md

   .

Work Phases

1. Confirm the review purpose, type, research questions, team roles, and registration requirements.
2. Predefine inclusion and exclusion criteria, information sources, complete search strategies, date ranges, and language ranges.
3. Finalize the protocol and search strategy before formal retrieval, and select the applicable registration platform. Use OSF for scoping reviews; do not submit to PROSPERO.
4. Verify immutable registration IDs, URLs, time-stamped with time zones, protocol versions, and hashes of registered files. Stop formal retrieval and reference import if registration requirements are not met.
5. Establish rules for deduplication, dual-review screening, conflict resolution, and exclusion justifications.
6. Determine extraction fields, bias assessment tools, and evaluation units for result layers based on study types.
7. Write a synthesis plan, heterogeneity handling, and sensitivity analysis before examining results.
8. Report protocol deviations, screening processes, study characteristics, biases, synthesized results, and limitations.

Operational Rules

• Select the review type first, then choose the reporting guideline and methodology manual.
• PRISMA is a reporting guideline, not a complete review implementation methodology.
• Scoping reviews are used to map concepts, evidence types, and gaps; they do not default to answering effect sizes.
• Inclusion and exclusion criteria for systematic reviews must be finalized before batch screening.
• Formal evidence synthesis must complete applicable protocol registration before formal retrieval; use OSF for scoping reviews.
• Protocol signing does not equal protocol registration. Do not proceed with retrieval without evidence of registration ID, URL, time-stamp, and frozen files.
• If the registered protocol or search strategy changes, stop and record the deviation; do not overwrite the registered version with new files.
• Search dates, platforms, complete search strategies, and result quantities must be retained.
• Rules for dual-review screening and conflict resolution must be determined before execution.
• Risk of bias assessment must match study design and specific results.
• Study quality, reporting completeness, and risk of bias cannot be combined into a single score.
• Whether to perform statistical pooling is determined by comparability and research questions, not by the number of studies.
• A narrative synthesis method must be predefined even if no meta-analysis is conducted.
• Record the time, reason, and impact on conclusions for protocol deviations.
• Evidence certainty assessment must be conducted at the result layer, with justifications for downgrading or upgrading provided.

Outputs

• Protocol, retrieval, and execution checklists.
• Registration packages and machine-readable registration records.
• Templates for screening, extraction, bias assessment, and synthesis.
• Deviation records, process data, and report checklists.

Stop Conditions

• Do not start batch screening if the research question and inclusion/exclusion criteria are not defined.
• Do not start formal retrieval, import, or screening if the review requiring registration has not completed registration.
• Stop and distinguish between trial retrieval and formal retrieval if the formal retrieval time is earlier than the registration time, or if the registered file hash does not match.
• Do not claim complete retrieval relying solely on one open database.
• Do not force pooling of studies if they are not comparable, even if conclusions are needed.
• Do not frame scoping reviews as conclusions on intervention effects.

Standalone Operation

• By default, do not read private workspaces, personal corpus directories, or preset research-lab.
• Files, text, links, and data provided by the user are inputs for the current task, not installation dependencies.
• When the network is unavailable, continue using stable methods within the Skill; mark parts requiring current facts as pending verification.
• Can connect with other research Skills in this package if they exist; if installed alone, directly return instructions for the next step without stopping the current task.

Source Discipline

• Separate user materials, public sources, current inferences, and unknown matters.
• Reporting guidelines only check report transparency and do not automatically prove design quality.
• Public source abstracts are stored within the Skill; return to official pages when requiring versions, fees, policies, APIs, or current journal requirements.
• Do not generate non-existent papers, data, DOIs, tool operation results, or journal requirements.
• Do not write user names, emails, institutions, accounts, tokens, absolute paths, or private project content into Skill, cases, tests, or public templates. Author information requiring disclosure on registration platforms is determined by the user on the platform and not written into Skill source code.