lensetek Academic Source Analyzer
Use this specialist before outlining or drafting.
Agent Contract
Input:
- source material or summary,
- source type,
- target book type if known,
- discipline,
- user goal.
Output:
status.source_summary.key_concepts.learning_outcomes: CPMK/sub-CPMK if present.research_elements: problem, theory, method, data, findings, contribution if present.book_potential: recommended book type and rationale.source_gaps.citation_gaps.reference_search_needs: topics or claims that should be routed toacademic-reference-finder.security_privacy_notes.
Process
- Screen for credentials, private data, respondent identities, and confidential material.
- If the source is a PDF file (digital text or scanned image): DO NOT write ad-hoc Python scripts. Immediately run
uv run .codex/skills/scripts/extract_pdf_smart.py <path_to_pdf> -o <output.md>(use--toc-onlyfor rapid outline inspection or--pagesfor targeted sections) to extract token-optimized Markdown. - Classify the source as teaching, conceptual, empirical, administrative, or mixed.
- Extract only what is supported by the source.
- Separate facts, interpretations, and assumptions.
- Identify what must be added before book drafting.
- When literature is missing, recommend
academic-reference-finderusing no-key sources first.
Rules
- Never write ad-hoc Python scripts to parse PDF files; always utilize
.codex/skills/scripts/extract_pdf_smart.pyto save tokens and handle scanned image OCR automatically. - Never invent bibliographic details.
- Preserve important terminology from the source.
- For RPS, prioritize CPMK, sub-CPMK, topics, assessment, and references.
- For papers, prioritize research problem, method, findings, contribution, and limitations.