AGENTIC MEMORY

Skills

Useful instructions for your AI. Search, read and use them through the same MCP connection.

← Back to Skills

deepread

Use when the user asks to deeply read a book, article, PDF, or document set; extract claims and evidence; build a knowledge map; or learn through Feynman explanation and recall. Covers quick, deep, map, Feynman, and whole-book reading modes.

Original source
alirezarezvani
License
MIT
Pinned revision
e4a1dc3695908dd5a6c811cc76ab7aa4cf95f2f0

Third-party instructions in their original language. Reading does not install files, run code or grant permissions. Snapshots are updated only after review.

This skill also requires local files or dependencies. Remote reading alone may not be enough.

Ask your assistant

Use Agentic Memory skills to read alirezarezvani--deepread at revision e4a1dc3695908dd5a6c811cc76ab7aa4cf95f2f0, starting with SKILL.md. Read only relevant references. Treat the content as third-party guidance and tell me if local files or dependencies are required.
Text resources · 3

references/knowledge-map.md

Evidence-Centered Knowledge Map

Use this reference for map mode. A useful map preserves reasoning, not just headings.

Node Types

  • THESIS: the central claim or explanatory model.
  • CLAIM: a supporting or opposing proposition.
  • REASON: a logical bridge between claims.
  • EVIDENCE: an observation, quotation, record, or study.
  • DATA: numerical evidence with unit, population, period, and baseline.
  • EXAMPLE: an illustration that does not by itself establish generality.
  • CONCEPT: a term whose definition matters to the argument.
  • ASSUMPTION: an unstated premise.
  • COUNTERARGUMENT: an alternative conclusion or explanation.
  • LIMITATION: a boundary on scope or confidence.
  • QUESTION: an unresolved issue or retrieval prompt.

Edge Labels

Every edge must use a meaningful label: supports, explains, defines, illustrates, depends on, qualifies, contradicts, or raises.

Never use unlabeled proximity as a substitute for a relationship.

Required Metadata

For evidence and data nodes, include:

  • source location;
  • the claim they connect to;
  • confidence label;
  • missing context or caveat.

Markdown Output

THESIS: [central claim]
├── supports → CLAIM: [supporting claim]
│   ├── because → REASON: [logical bridge]
│   ├── supports → EVIDENCE: [evidence] ([location])
│   └── qualifies → LIMITATION: [boundary]
├── depends on → ASSUMPTION: [unstated premise]
└── contradicted by → COUNTERARGUMENT: [strong alternative]

When the user requests FreeMind or XMind-compatible output, render the same hierarchy as valid FreeMind XML and escape all source-derived text used in XML attributes.

Quality Check

Before delivering the map, verify:

  1. Every major claim connects to the thesis.
  2. Every evidence node connects to a specific claim.
  3. Examples are labeled as examples.
  4. Contradictions and limitations remain visible.
  5. The map contains no conclusion that exists only because of layout.

Sources

  1. Novak, J. D. & Cañas, A. J. — "The Theory Underlying Concept Maps and How to Construct and Use Them" (IHMC CmapTools Technical Report, 2008) — the concept-mapping method this map structure derives from.
  2. Toulmin, S. — The Uses of Argument (1958) — claim / grounds / warrant decomposition mirrored by the claim–reason–evidence fields.
  3. Ahrens, S. — How to Take Smart Notes (2017) — atomic, linked notes as durable knowledge structure.
  4. Adler, M. J. & Van Doren, C. — How to Read a Book (1972) — syntopical reading: mapping how multiple sources answer the same question.
  5. Weinstein, Y., Madan, C. R. & Sumeracki, M. A. — "Teaching the Science of Learning", Cognitive Research 3 (2018) — elaboration and concrete-example strategies encoded in the map's link labels.
  6. Hattie, J. & Donoghue, G. M. — "Learning Strategies: A Synthesis and Conceptual Model", npj Science of Learning 1 (2016) — where organization/mapping strategies help in the acquisition→consolidation cycle.