| name | dh-explore |
| description | Digital humanities research discovery — research question development, digital archive search, primary source criticism, corpus assembly, literature review, and research framework mapping. Produces venue-specific framework diagrams showing recommended methods (topic modeling, NER, sentiment analysis, close reading) and paper structure for NLP/AI venues, DH journals, traditional humanities journals, or specific target publishers. Use when starting a DH project, building a corpus, or planning a research methodology. |
| metadata | {"version":"1.1.0","last_updated":"2026-05-29","status":"active","data_access_level":"redacted","task_type":"open-ended","related_skills":["dh-write","dh-refs","dh-pipeline"]} |
dh-explore — Research & Discovery
What This Skill Does
dh-explore guides you from an initial DH idea to a fully developed research framework, including:
- A refined, dual-framed research question (interpretive claim + methodological contribution)
- A methodology blueprint matched to your corpus and question
- A Corpus Manifest with primary sources assessed for OCR quality and provenance
- An annotated bibliography spanning both humanities and computational DH literature
- A Research Question Brief ready to hand off to
dh-write
Who It's For
- Researchers at any stage — undergraduate to faculty
- Projects anywhere on the DH methodology spectrum (close reading to large-scale text mining)
- Particularly strong for projects involving digitized historical materials and noisy OCR text
Modes and Trigger Keywords
| Mode | Trigger | What Happens |
|---|
full | /dh-explore or /dh-explore full | Complete pipeline: intake → socratic → methodology → archives → literature → corpus → synthesis → RQ → framework |
quick-brief | /dh-explore quick-brief | Fast path: intake → RQ refinement only; ~15 min |
socratic | /dh-explore socratic | Guided dialogue to discover your research question |
archive-survey | /dh-explore archive-survey | Archive identification and evaluation only |
methodology-guide | /dh-explore methodology-guide | Method selection and OCR quality assessment only |
lit-review-humanities | /dh-explore lit-review-humanities | Humanities literature search and annotated bibliography |
lit-review-technical | /dh-explore lit-review-technical | Computational DH literature review |
framework | /dh-explore framework | Research framework map: given an RQ, build method diagram + venue-specific paper structure |
Mode Routing
When this skill is invoked, do the following:
-
Always start with intake: Read agents/intake_agent.md and execute it to configure the session.
-
After intake, route by mode:
full → Read and execute agents in order: socratic_mentor_agent.md → methodology_advisor_agent.md → digital_archive_agent.md → primary_source_agent.md → corpus_builder_agent.md → humanities_literature_agent.md → technical_literature_agent.md → cross_domain_synthesis_agent.md → research_question_agent.md → framework_architect_agent.md
quick-brief → Read and execute: research_question_agent.md only (after intake)
socratic → Read and execute: socratic_mentor_agent.md → research_question_agent.md
archive-survey → Read and execute: digital_archive_agent.md → primary_source_agent.md → corpus_builder_agent.md
methodology-guide → Read and execute: methodology_advisor_agent.md
lit-review-humanities → Read and execute: humanities_literature_agent.md
lit-review-technical → Read and execute: technical_literature_agent.md
framework → Read and execute: framework_architect_agent.md (requires RQ — ask if not present; can run standalone or after any other mode)
-
If upstream materials are detected (blocks labeled ## Research Question Brief, ## Corpus Manifest, ## Methodology Blueprint): skip agents that would produce what's already present; use the existing materials as input to agents further down the chain.
Shared References
Before beginning, load these shared resources:
- Read:
../shared/references/dh_methodology_spectrum.md
- Read:
../shared/references/intent_clarification_protocol.md
- Read:
../shared/references/anti_leakage_protocol.md
Handoff Output
On completion of full mode, produce a dh-explore Output Package containing:
## Research Question Brief ← for dh-write intake
## Methodology Blueprint ← for dh-write methodology_writer_agent
## Corpus Manifest ← for dh-write and dh-pipeline
## Literature Review Brief — Humanities
## Literature Review Brief — Technical
## Research Framework
## Research Framework Map ← NEW: Mermaid diagram + venue-specific paper structure
dh-write's intake agent detects this package and auto-populates the writing session. Save or paste this package when starting /dh-write.
The Canonical Use Case
This skill was built around a project working with 19th-century abolitionist newspapers from Chronicling America — digitized historical text with 75–85% OCR accuracy. If your project involves:
- Noisy OCR text from historical digitization
- Primary source provenance questions
- The tension between computational pattern-finding and historical interpretation
...then you are using this skill for exactly the purpose it was designed for.
Exemplar DH Projects Referenced in This Skill
- "Mining for the Meanings of a Murder: The Impact of OCR Quality on the Use of Digitized Historical Newspapers" — DHQ — OCR quality and text mining methodology
- "Machine Reading the Primeros Libros" — DHQ vol.10 no.4 — Colonial archive digitization
- "How Network Analysis Uncovers International Networks of Smuggling History" — Cultural Analytics 2023 — Historical network analysis
- "The Generative Dissensus of Reading the Feminist Novel, 1995-2020" — Cultural Analytics — Computational + close reading synthesis
- "Slow, Painful and Expensive: Current Challenges in Text-Mining Corpus Construction" — DHQ 19/3 — Corpus construction reality