Scholarly Literature Review — policy
Purpose
Define the normative rules for the Scholarly Literature Review Stack.
This policy applies to scholarly workflows that produce one of the following outputs:
- literature review
- evidence synthesis
- source comparison
- gap mapping
- annotated bibliography
- evidence table
The policy sets the evidence boundary, source-grounding requirements, reporting discipline, and the conditions under which formal review labels may or may not be used.
Canonical links
- System prompt template (copy/paste): scholarly-evidence-gate.system.txt
- Workflow prompt template: literature-review.user.txt
- Workflow prompt template: evidence-synthesis.user.txt
- Workflow prompt template: source-comparison.user.txt
- Workflow prompt template: gap-mapping.user.txt
- Workflow prompt template: annotated-bibliography.user.txt
- Workflow prompt template: evidence-table.user.txt
- Base policy (normative): Facts-only: Authoritative sources required (citations required)
- Base policy (normative): Web Verification & Citations Policy
- Prompt templates index: Prompt templates
Non-negotiable rules (normative)
S1) Base-policy inheritance (HARD)
This policy MUST comply with all normative rules in:
S2) Source-grounding (HARD)
- Material factual claims MUST be grounded in sources that were actually inspected in the current run.
- Claims about user-provided files, logs, datasets, repositories, or ZIP archives require the relevant artifact to be available and inspected in the current conversation.
- Unsupported inference, guesswork, or pattern-completion MUST NOT be presented as fact.
S3) Source preference (HARD)
- Prefer authoritative, directly relevant, and original sources whenever possible.
- When a claim is based on scholarly literature, prefer direct references to original research sources over indirect summaries when the original source is available.
- If a preprint is cited, it MUST be identified as a preprint.
- If a peer-reviewed published version exists for a cited preprint and is available, prefer the published version.
S4) Workflow scope (HARD)
This policy governs scholarly review outputs limited to:
- literature review
- evidence synthesis
- source comparison
- gap mapping
- annotated bibliography
- evidence table
The stack MUST NOT be used to imply methods or reporting standards that were not actually followed.
S5) Formal review labels (HARD)
- Do NOT describe an output as a systematic review, PRISMA-compliant review, scoping review compliant with a formal reporting standard, or protocol-based review unless the user explicitly requests that review type and provides the method/reporting detail required to support that label.
- PRISMA 2020 is not a default label for general literature review outputs.
- PRISMA-ScR is not a default label for gap mapping or evidence-map outputs.
S6) Search and source transparency (HARD)
When the workflow uses external search or browsing, the output MUST state:
- which sources or source locations were actually used
- the effective scope or limits applied to source selection
- any material evidence limits caused by the available source set
If search details are unavailable, the output MUST NOT imply more rigorous search coverage than was actually performed.
S7) Evidence handling (HARD)
- Included-source findings MUST be kept separate from synthesis, interpretation, or comparison summary.
- If sources disagree, the disagreement MUST be stated explicitly.
- If the evidence is limited, mixed, weak, outdated, or incomplete, that limitation MUST be made visible in the output.
- The output MUST stay within what the inspected sources support.
S8) Output discipline (HARD)
- Output MUST use clear, formal, technical scholarly register.
- Output MUST separate supported findings from uncertainty, unresolved conflict, and evidence gaps.
- Output MUST avoid rhetorical or inflated claims about certainty, completeness, or rigor.
Related indexes
- Policies: /policies/
- How-to: /how-to/
- Prompt templates: /prompts/