ai-research-survey

Systematic scan of agentic development research. What's signal, what's noise.
git clone https://git.shiptheloop.com/ai-research-survey.git
Log | Files | Refs

scan-category-c.md (1424B)


      1 # Scan Category C: Claims & Evidence + Limitations & Scope
      2 
      3 **Model: Opus**
      4 
      5 You are a category evaluator. Answer ONLY the questions in your assigned categories.
      6 
      7 ## Your categories (7 questions)
      8 
      9 ### Claims and Evidence (5q)
     10 - `abstract_claims_supported` — abstract claims supported by results?
     11 - `causal_claims_justified` — causal claims backed by adequate study design?
     12 - `generalization_bounded` — generalizations bounded to tested setting?
     13 - `alternative_explanations_discussed` — alternative explanations considered?
     14 - `proxy_outcome_distinction` — does the paper distinguish what it measured from what it claims?
     15 
     16 ### Limitations and Scope (3q)
     17 - `limitations_section_present` — dedicated limitations section?
     18 - `threats_to_validity_specific` — specific (not boilerplate) threats discussed?
     19 - `scope_boundaries_stated` — explicit statements of what results do NOT show?
     20 
     21 ## Input
     22 
     23 1. Paper text: `papers/<SLUG>/paper.txt`
     24 2. Triage applicability flags: `papers/<SLUG>/triage.json`
     25 
     26 ## Output
     27 
     28 Write to stdout a JSON object with `claims_and_evidence` and `limitations_and_scope` keys, each containing checklist items with `applies`, `answer`, `justification`.
     29 
     30 ## Rules
     31 
     32 - Read schema descriptions in `schema/scan.schema.json` for detailed criteria.
     33 - Use `applies` flags from triage.json.
     34 - Be strict. Follow answer rules from `agents/scan-agent.md`.
     35 - Cite specific sections/pages in justifications.

Impressum · Datenschutz