How to use the Strategic Reading agent skill

What the strategic-reading skill makes the agent do

You give the agent a long source plus a specific strategic problem, such as a negotiation, a competitive fight or a difficult decision. The agent reads the source with that problem in mind, maps the source's central ideas onto your situation, and produces a structured analysis.

The skill is explicitly not for general book summaries. It works only on the text you supply, so it cannot check current news or verify up-to-date facts.

• Maps the source's central dynamic onto your situation
• Triages chapters or sections as HIGH, MEDIUM or LOW relevance
• Lists what to do, what to avoid and what to watch for, citing the source
• Gives short-term (this week or month), medium-term (this quarter) and long-term (this year+) recommendations
• Selects direct quotes from the source, up to 5 to 10

Install the strategic-reading skill

The skill is distributed as a ZIP file from the directory page. You can install it in one of two ways. Choose the one that matches how you use your agent.

Option A is a personal install that makes the skill available everywhere on your machine. Download the ZIP from the page and unzip it into ~/.claude/skills/. Option B is a project install that keeps the skill inside one project. Unzip it into the .claude/skills/ folder of that project.

Option C is to upload the ZIP in the Claude apps under Settings → Capabilities → Skills.

• Personal: unzip into ~/.claude/skills/
• Project: unzip into .claude/skills/ inside your project folder
• Claude apps: Settings → Capabilities → Skills, then upload the ZIP

Check what the skill package contains

The package's main file is SKILL.md. It holds the skill's name, trigger phrases and the full process the agent follows. Reading it is the best way to see exactly what the agent will do before you rely on it.

• SKILL.md: frontmatter (name, description, triggers) and the full instructions

Trigger the skill with natural requests

You do not need a special command. The skill responds to phrases close to its listed triggers. Include the source and the problem in the same message so the agent knows what to read and what to analyze it against.

• "Strategic reading: here is the case study on the 2008 supplier dispute. My problem is that our largest client wants a 20% price cut."
• "Read this through the lens of our market-entry problem: [paste transcript or attach EPUB]."
• "Apply this to my problem: I'm negotiating with a gatekeeper who controls distribution."
• "What can I learn from this about handling a hostile board?"
• "Extract a playbook from this article about competitor pricing, for my pricing decision."

Understand the output and where it is saved

The result is a markdown page written to your knowledge base, not a chat-only summary. Pages are filed by subject. Problem-specific playbooks go to projects/<slug>/playbook.md, and general strategy pages go to concepts/<slug>.md. The page includes an executive summary, the core parallel, a chapter triage, the source's playbook, counter-tactics, the applied playbook, key quotes and see-also links.

A PDF version is optional. The skill treats the markdown page as the main artifact and PDF only as a rendering.

• Core Parallel, Chapter Triage, Source's Playbook, Counter-Tactics, Applied Playbook, Key Quotes, See Also
• Saved to projects/ for problem-specific work or concepts/ for general strategy

Know what the skill needs and what it runs

The skill needs a knowledge base or brain setup that accepts put_page writes, because it writes markdown pages to concepts/ or projects/. You must also supply the source text yourself, in EPUB, PDF, article or transcript form.

Some tools are optional. The process references pdftotext for PDFs and BeautifulSoup for EPUBs. The brain-pdf skill is needed only if you want a PDF rendering.

According to the skill's metadata, it runs no scripts and uses no network access. Its contract is limited to writing pages under the listed directories.

• Required: a brain setup that accepts put_page writes
• Required: source text you provide (EPUB, PDF, article or transcript)
• Optional: pdftotext for PDFs, BeautifulSoup for EPUBs
• Optional: the brain-pdf skill for a PDF rendering
• Runs scripts: no. Uses network: no.

Safety notes on scripts and network access

Because the skill writes files, it is marked as mutating. It writes pages into the concepts/ and projects/ folders of your knowledge base. Check those folders after a run if you care where content lands.

The skill's metadata says it runs no scripts and uses no network. Some sections of the file show illustrative fetch or search examples, and the skill says these are not API calls. The skill's own inputs are the text you upload and the brain pages it writes.

• Writes markdown pages to concepts/ and projects/ only
• Does not run scripts or make network calls, per its metadata
• Works on your supplied text, so it will not add outside facts

Tips for better playbooks

Describe the problem concretely. A specific situation, with the parties and the decision you face, gives the agent a sharper lens than a vague topic. Supply the full text or the relevant chapters, since the agent can only use what it reads.

If the agent reports that part of the source was unavailable or truncated, it should say which parts it read and should not extend the playbook beyond them. Check the Key Quotes against the source before acting on them.

• State the problem, the other parties and the decision in one or two sentences
• Provide complete chapters or a full transcript where possible
• Ask for a PDF only after the markdown page looks right