RW Evidence Map

Initiation

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 is needed, 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, open the official links in

   references/source-map.md

   for verification and record the date.

Work Phases

1. Define the research questions, decisions, and time scope for the evidence map service.
2. Establish a stable ID for each study and link it to the original text or verifiable identifiers.
3. Extract information on study design, population, exposure or intervention, comparison, outcomes, timing, and effects.
4. Record risk of bias, applicability, precision, and directness separately.
5. Label relationships as supporting, refuting, limiting, incomparable, or pending verification.
6. Output evidence clusters, causes of conflicts, gaps, and next steps for evidence collection.

Operational Rules

• The evidence map first fixes node definitions before establishing relationships.
• Study conclusions and current comprehensive judgments must be stored separately.
• Different constructs, measurement tools, or time points cannot be merged just because their names are similar.
• Effect direction, effect size, and uncertainty must be recorded separately.
• Lack of statistical significance does not mean no effect exists.
• Studies with high risk of bias can be retained but cannot be treated equally with low-risk studies.
• When conflicts arise, first check for differences in population, design, measurement, analysis, and timing.
• Evidence gaps should be distinguished between no research available, no full text accessible, and unreported results.
• Each strict relationship must be supported by the original text location or structured records.
• Evidence certainty is a judgment at the outcome level, not a score assigned to the entire paper.
• The map can display relationships but cannot automatically complete causal inference.
• When adding new evidence, retain the old judgments and reasons for changes.

Output

• Evidence matrix or relationship diagram.
• Records of relationship sources, biases, and comparability.
• Conflict explanations, gaps, and next steps.

Stop Conditions

• When the abstract is insufficient to judge the method or results, mark it as pending full-text verification.
• Do not perform numerical merging when there is a lack of common outcome standards.
• Do not treat the number of connections in the map as evidence strength.
• Relationships without source localization can only enter the candidate area.

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 input 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; when installed separately, directly return instructions for the next step without stopping the current task.

Source Discipline

• Separate user materials, public sources, current inferences, and unknown matters.
• Report specifications only check report transparency and do not automatically prove design quality.
• Public source abstracts are stored within the Skill; return to official pages when information on versions, fees, policies, APIs, or journal status is needed.
• Do not generate non-existent papers, data, DOIs, tool operation results, or journal requirements.