Skip to content

R0049/2026-03-31/Q001 — Assessment

BLUF

No published AI/LLM system prompt was found that implements a complete analytical rigor framework for research. Partial implementations exist — notably Scott Roberts' LLM-based structured analytic techniques and the Framework Chain-of-Thought prompting approach — but these address individual techniques or single research phases rather than end-to-end methodology. The gap between what exists and a comprehensive framework is substantial.

Probability

Rating Very unlikely (05-20%) that a complete, published framework prompt exists and was not found
Confidence Medium-High
Confidence rationale Multiple independent search strategies across academic databases, GitHub, and prompt libraries converged on the same finding. The search covered the most likely publication venues. However, unpublished corporate/government prompts and non-English-language publications represent potential blind spots.

Reasoning Chain

  1. The query asks whether a published prompt exists that implements a full analytical rigor framework for AI research.
  2. Three independent search strategies were executed: academic literature (S01), GitHub repositories (S02), and practitioner publications (S03).
  3. Academic literature revealed framework extensions (PRISMA-trAIce SRC01-E01, L-PRISMA) and methodology enhancements (Framework CoT SRC02-E01) — but these are reporting checklists and single-phase prompts, not complete system prompts.
  4. GitHub repositories revealed the largest prompt libraries (SRC03-E01) lack research methodology prompts entirely, and research automation tools (SRC05-E01) lack analytical rigor mechanisms.
  5. Practitioner publications revealed Roberts' SAT implementations (SRC04-E01) — the closest match — but these are three separate tools, not a unified framework.
  6. The pattern across all sources is consistent: the field has invested in automating research tasks but not in encoding analytical discipline.

Evidence Base Summary

Source Reliability Relevance Key finding
SRC01 High Medium Reporting checklist, not operational prompt
SRC02 Medium-High High Framework-guided prompting works (93.6% accuracy) but covers screening only
SRC03 Medium Medium Largest prompt libraries lack research methodology content
SRC04 Medium High Individual SATs implemented as LLM tools, not unified framework
SRC05 Medium-High Medium Research automation without analytical rigor

Collection Synthesis

Dimension Assessment
Evidence quality Mix of peer-reviewed (SRC01), preprint (SRC02), and practitioner (SRC04) sources — adequate for landscape characterization
Source agreement High agreement across all sources: partial implementations exist, comprehensive frameworks do not
Independence Sources are independent — different authors, venues, and approaches converge on same finding
Outliers None — all evidence points in the same direction

Detail

The evidence collection reveals a consistent landscape: the AI research community has produced (a) reporting frameworks for AI use in research (PRISMA-trAIce), (b) single-technique prompt implementations (Framework CoT, Roberts' SATs), and (c) automated research workflow platforms (Agent Laboratory, STORM, PaperQA2). None of these combine into a comprehensive analytical rigor system prompt. The gap exists at the integration layer — combining multiple techniques into a unified methodology.

Gaps

Gap Impact on confidence
Unpublished corporate/government prompts not searchable Could reduce confidence if classified or proprietary frameworks exist
Non-English language publications not searched Minor — framework development appears concentrated in English-language venues
Preprint status of key source (SRC02) Minor — findings may change but core landscape characterization is stable
Internal AI company research (Anthropic, OpenAI, Google) Could reduce confidence — these organizations may have internal methodology prompts

Researcher Bias Check

The researcher is building a unified research methodology system, which creates an incentive to find that no prior art exists. This bias was mitigated by: (a) designing searches specifically to find confirming evidence for H1, (b) evaluating partial implementations generously, and (c) documenting Roberts' SAT work as genuine prior art rather than dismissing it.

Cross-References

  • Q002 — Related: whether IC and scientific frameworks have been combined at all (for any use)
  • Q003 — Related: whether tools (not prompts) implement structured analytical frameworks