Maintenance in progress: we are indexing a large batch of new skills. Some pages may load slowly or briefly show no results. Nothing is lost, and everything is back to normal within the hour.
Use this skill when the situation clearly matches its intended use cases
Combine with related skills for comprehensive analysis
Iterate on outputs if initial results don't fully meet requirements
Common variations:
Adjust the depth of analysis based on available time and information
Scale the approach for different levels of complexity
Adapt the output format to audience needs
When to skip this skill:
The situation doesn't match the core use cases
Simpler approaches would be more appropriate
Time constraints require faster methods
Example
User: "I'm researching a company's founding story for a history project. I've read the founder's biography and some news articles."
Turn Every Page response:
You've barely started. Let me map what's actually available.
Immediate gaps in your research:
Corporate filings - Articles of incorporation, SEC filings, annual reports. These aren't glamorous but they reveal dates, partners, and money that biographies gloss over.
The people who left - Every company has early employees who departed. They have stories the official history doesn't tell. Who were the first 10 hires? Where are they now?
Local newspaper archives - Not just the business section. Check obituaries of early associates, society pages, real estate transactions.
The competitors - What did rivals say about this company at the time? Trade publications, industry newsletters?
Court records - Any lawsuits, even ones settled? These reveal conflicts the biography won't mention.
The founder's papers - Are they archived somewhere? Universities, historical societies, company archives?
Your news articles are tertiary sources—reporters summarizing what they were told. The founder's biography is what he wanted remembered.
The real story is in:
What he didn't put in the biography
What the people who left would tell you
What the court records show
What the financial filings reveal
What archive access do you have? What court jurisdiction would have relevant records?
Integration
This skill pairs with:
silence-interview-technique - After documents, interview people
power-analysis-lens - Documents reveal power mechanisms
This is research methodology, not actual database access
Cannot access proprietary archives or classified documents
Time and feasibility constraints are real—acknowledge them
Not every project justifies Caro-level research (decades)
Source Expert
Robert Caro - experts/robert-caro/
1---2name: turn-every-page-research3description: Exhaustive document research methodology that ensures no source is overlooked. Based on Robert Caro's foundational research principle.4license: MIT5---67# Turn Every Page Research89Exhaustive document research methodology that ensures no source is overlooked. Based on Robert Caro's foundational research principle.1011---1213## When to Use1415- Research feels shallow or incomplete16- You're uncertain if you've found everything relevant17- Writing about historical events, organizations, or people18- Investigating a topic where official accounts may be incomplete19- Preparing for in-depth biographical or investigative work2021**Trigger Phrases:**22- "My research feels shallow"23- "Am I missing something?"24- "How do I research more thoroughly?"25- "What sources should I check?"26- "I need to dig deeper"2728---2930## Inputs3132| Input | Required | Description |33|-------|----------|-------------|34| topic | Yes | The subject being researched |35| sources_checked | No | List of sources already consulted |36| research_goal | No | What you're trying to discover or prove |37| time_available | No | Constraints on research depth |3839---4041## Core Principle4243> "Turn every page. Never assume anything. The next page might have what you're looking for."44> — Robert Caro4546The breakthrough is always in the document everyone else skipped. Don't assume relevance before reading. Don't trust summaries—read originals.4748---4950## Workflow5152### Step 1: Identify All Possible Sources5354Before reading anything, map the complete universe of sources:5556**Primary Sources:**57- [ ] Official records, filings, registrations58- [ ] Internal documents, memos, correspondence59- [ ] Meeting minutes, transcripts, proceedings60- [ ] Financial records, contracts, agreements61- [ ] Personal papers, diaries, letters62- [ ] Photographs, recordings, artifacts6364**Secondary Sources:**65- [ ] Published biographies, histories, studies66- [ ] Newspaper archives from the relevant period67- [ ] Magazine and journal articles68- [ ] Documentaries, interviews, oral histories69- [ ] Academic theses and dissertations7071**Overlooked Sources:**72- [ ] Footnotes and bibliographies in existing works73- [ ] Assistants, secretaries, and junior staff74- [ ] Contemporaries who were rivals or critics75- [ ] Regional and local archives76- [ ] Collections of related figures7778### Step 2: Create Reading Protocol7980For each archive or document collection:81821. **Read everything** - Not just relevant-seeming documents832. **Take notes on every page** - Even if "irrelevant"843. **Note what's missing** - Gaps reveal hidden stories854. **Cross-reference** - Documents illuminate each other865. **Return later** - After more context, reread with fresh eyes8788### Step 3: Apply the "Never Assume" Rule8990| Instead of... | Do this... |91|---------------|------------|92| Skipping dense legal documents | Read every clause |93| Trusting published summaries | Find original documents |94| Assuming a source is irrelevant | Read it anyway |95| Stopping when you find "enough" | Ask: What else exists? |96| Believing official versions | Look for contradictions |9798### Step 4: Document Your Research Path99100Track:101- What archives/collections exist102- What you've read vs. what remains103- Promising leads to follow104- Questions that emerged105- Contradictions discovered106107### Step 5: Identify Gaps and Pursue108109The most important discoveries come from:110- Documents that should exist but don't111- People who were there but never interviewed112- Archives no one else has accessed113- The footnote in another researcher's work114115---116117## Outputs118119| Output | Description |120|--------|-------------|121| Source inventory | Complete list of potential sources |122| Reading log | What has been read, what remains |123| Discovery notes | Key findings, contradictions, leads |124| Gap analysis | What's missing that should exist |125| Research plan | Prioritized next steps |126127---128129130131**Format:**132```markdown133## Analysis: [Topic]134135### Key Findings136- [Finding 1]137- [Finding 2]138- [Finding 3]139140### Recommendations1411. [Action 1]1422. [Action 2]1433. [Action 3]144```145146147## Additional Notes148149**Best practices:**150- Use this skill when the situation clearly matches its intended use cases151- Combine with related skills for comprehensive analysis152- Iterate on outputs if initial results don't fully meet requirements153154**Common variations:**155- Adjust the depth of analysis based on available time and information156- Scale the approach for different levels of complexity157- Adapt the output format to audience needs158159**When to skip this skill:**160- The situation doesn't match the core use cases161- Simpler approaches would be more appropriate162- Time constraints require faster methods163164## Example165166**User:** "I'm researching a company's founding story for a history project. I've read the founder's biography and some news articles."167168**Turn Every Page response:**169170You've barely started. Let me map what's actually available.171172**Immediate gaps in your research:**1731741. **Corporate filings** - Articles of incorporation, SEC filings, annual reports. These aren't glamorous but they reveal dates, partners, and money that biographies gloss over.1751762. **The people who left** - Every company has early employees who departed. They have stories the official history doesn't tell. Who were the first 10 hires? Where are they now?1771783. **Local newspaper archives** - Not just the business section. Check obituaries of early associates, society pages, real estate transactions.1791804. **The competitors** - What did rivals say about this company at the time? Trade publications, industry newsletters?1811825. **Court records** - Any lawsuits, even ones settled? These reveal conflicts the biography won't mention.1831846. **The founder's papers** - Are they archived somewhere? Universities, historical societies, company archives?185186Your news articles are tertiary sources—reporters summarizing what they were told. The founder's biography is what he wanted remembered.187188**The real story is in:**189- What he didn't put in the biography190- What the people who left would tell you191- What the court records show192- What the financial filings reveal193194What archive access do you have? What court jurisdiction would have relevant records?195196---197198## Integration199200This skill pairs with:201- **silence-interview-technique** - After documents, interview people202- **power-analysis-lens** - Documents reveal power mechanisms203- **scene-over-summary** - Rich sources enable vivid scenes204205---206207## Constraints208209- This is research methodology, not actual database access210- Cannot access proprietary archives or classified documents211- Time and feasibility constraints are real—acknowledge them212- Not every project justifies Caro-level research (decades)213214---215216## Source Expert217218Robert Caro - `experts/robert-caro/`
Run npx skillmds add sethmblack/turn-every-page-research in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.
Exhaustive document research methodology that ensures no source is overlooked. Based on Robert Caro's foundational research principle. It is listed under Research & Search on SkillMD.
This skill has not completed SkillMD's automated safety review yet. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.
This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.
Yes. Installing skills from SkillMD is free. This skill is licensed under MIT.
sethmblack (@sethmblack) published this skill. Their other Agent Skills are listed on their SkillMD profile.