Read a source through the lens of a specific problem. Produces an applied playbook, not a summary.
Strategic Reading
Read a book, article, paper, transcript, or case study through the lens of a specific strategic problem. Produces an applied playbook that maps the source's insights onto the problem and gives actionable recommendations.
This is not summarization. This is reading with a mission.
Contract
- Requires TWO inputs: source text AND a specific strategic problem or question
- Output is an applied analysis, not a general summary
- Every recommendation has its own passage locator, source claim, and separately labeled applied inference
- Never runs on a source without a clear "what problem does this help with?"
Evidence for each recommendation
Give every proposed tactic/action a stable ID and its own evidence record:
- Passage locator: identify the source edition/revision or saved snapshot and the exact page/section/paragraph, timestamp, or anchored text range. A title, homepage, or quote shared across a whole section is insufficient. If the text has no locators, preserve an authorized excerpt and assign paragraph IDs in that excerpt; label them as assigned IDs, not original page numbers. Never invent a page or imply an unseen passage was read.
- Source claim: quote briefly or faithfully paraphrase what that passage actually establishes, including its scope and qualifications.
- Applied inference: state the proposed action for the user's problem and explain the bridge from the source claim. Label this as your application, not something the source prescribed for the user.
- Assumptions and limits: identify what must hold for that bridge to work and how the user could check it. Keep an unsupported idea explicitly labeled as a hypothesis outside the source-backed recommendation list.
Repeat the locator/claim/inference record for each recommendation, even if two use the same passage. A section-level quote never covers untraced extra tactics. Use recommendation IDs in a timing list instead of adding fresh ungrounded actions there. Distinguish evidence-backed relevance from a guaranteed result. See traceable-example.md for two independently traceable recommendations and a fully available fictional source.
When to use
- You have a strategic problem and a source that might inform it
- An academic paper or book needs to be read for actionable takeaways
- A career, research, or project decision could be informed by external knowledge
- Learning with intent: "I need to understand this SO I can DO that"
When NOT to use
- Casual reading or general interest — that's a book note, not strategic reading
- Pure fact-finding — use a research skill instead
- First-pass source familiarization — read the source first, then apply
Phases
- Identify the strategic problem. What specific situation or decision is the user working on? Get this explicit before reading the source.
- Read the source. Extract key frameworks, tactics, observations, and historical patterns.
- Map to the problem. Triage sections for relevance (HIGH / MEDIUM / LOW), with passage locators and accurate source claims. Keep application distinct.
- Build the playbook. Give each recommendation the evidence record above, then organize its ID by timing.
- Save if authorized. Write to the explicitly chosen project/reading destination only when the task authorizes it; otherwise return the playbook in the conversation.
Output Structure
---
title: "{Source Title} — Applied to {Problem}"
type: applied-reading
date: YYYY-MM-DD
source: "{citation}"
problem: "{the strategic question}"
---
# {Source Title} — Applied to {Problem}
> Brief overview: the problem, the relevant source claims, and the proposed application.
## The Core Parallel
How the source's central dynamic maps onto the user's situation.
## Section Triage
For each major section of the source:
- 2-3 sentence summary of what it says
- Relevance to the problem: HIGH / MEDIUM / LOW
- One directly applicable quote or faithful paraphrase with its passage locator
## The Playbook
Repeat for EACH recommendation, including tactics and proposed avoidances:
### R1 — {proposed action}
- **Passage locator:** {source identity/revision + precise location}
- **Source claim:** {what the passage actually says}
- **Applied inference:** {action + why it may fit this problem}
- **Assumptions/limits:** {conditions, uncertainty, and a check}
## Short/Medium/Long-Term Actions
- **Now:** recommendation IDs from the playbook
- **This month:** recommendation IDs from the playbook
- **This quarter:** recommendation IDs from the playbook
## Connections
Related notes, projects, and concepts this analysis touches.
Anti-Patterns
- Producing a generic book summary instead of applied analysis
- Reading without a clear strategic question — get the problem first
- Forgetting to map recommendations back to the source passages
- Writing analysis disconnected from the user's actual situation
- Over-recommending: 2-3 high-impact moves beat 10 low-impact suggestions
Related skills
- Downloads
- 703All time
- Security audit
- Pass
- Last updated
- 1w ago
- Current version
- v1.0.2
- License
- MIT-0
