passage-exegesis

Passage Exegesis Skill

Systematic exegetical analysis of a Bible passage using all available TheologAI tools.

When to Use

Use this workflow when the user asks for:

• Exegesis or exposition of a passage
• Deep study of a Bible text
• Understanding what a passage means
• Sermon preparation help
• Theological analysis of a text

Methodology

Step 1: Establish the Text

• Call bible_lookup with the reference in the user's preferred translation (default: ESV)
• Also retrieve KJV for a second perspective
• If NET is available, retrieve it too for the translator notes
• Read the passage carefully before analyzing

Step 2: Original Language Analysis

• Trace the passage's literary and discourse flow before selecting terms.
• For a range, select at most three key individual verses. Never pass a range to bible_verse_morphology or original_language_study.
• Study only terms affecting an actual translation or interpretive question.
• Call original_language_study with one verse and the verse-local target. Resolve returned ambiguity by source position, not guesswork.
• Request usage_level: "overview" from an exact original_language_lookup only after the verse-local analysis and only when global distribution is consequential. Overview is totals plus complete books only. Do not request global usage from original_language_study.
• Treat morphology as constraining evidence, not a self-sufficient interpretation.

Step 3: Literary Context

• Start with parallel_passages using exactly reference, corpora: ["ubs_source_attested"], maxGroups: 5, and includeText: false. Browse returned structured source-attested metadata only; do not flatten group membership into inferred pairwise relationships.
• Preserve every returned group's member order, source order, and provenance. The source labels directionality unspecified, so never infer quotation, dependence, synoptic direction, or a thematic relationship.
• Continue only when sourceAttestedResultWindow.additionalMatchStatus is additional_match_observed, it includes nextCursor, and another page could materially help. Pass nextCursor unchanged as groupCursor while preserving exactly the same reference, corpora, and maxGroups. Omit includeText, translation, includeAlignment, legacy controls, and OpenBible controls. Never inspect, decode, or rewrite the cursor.
• Stop when the result window is terminal, the evidence is sufficient, or after two continuation calls (at most three pages / 15 groups). If the cap ends a potentially useful survey, state that the survey was bounded.
• From the browsed metadata, select at most two materially relevant groups in returned source order. Preserve each selected group and every member intact; do not dedupe or flatten the groups. Only when comparison text materially helps, construct a separate text-enrichment queue by traversing selected groups and their members in returned source order. Dedupe only that queue by normalizedReference, keeping its first occurrence. Use direct bible_lookup only for the first 12 unique queue references; label every later unique queue reference budget_omitted before lookup. Record lookup failures and budget omissions; never backfill with later references. Never combine groupCursor with includeText.
• Use corpora: ["theologai_legacy"] only when the user specifically wants the older curated relationship labels (synoptic, quotation, allusion, or thematic).
• Request includeOpenBibleCrossReferences: true only for a separate, broader cross-reference list; do not present those rows as UBS-attested group members.
• Call bible_cross_references to identify broader OpenBible.info discovery leads across Scripture. Prefer its structured requested/resolved references, positioned rows, threshold-scoped result window, and provenance; preserve raw vote order. Keep this evidence separate from UBS groups, and do not infer a relationship classification or directionality.
• Treat those links as candidates for contextual investigation, not proof of quotation, literary dependence, a source-attested parallel, or even a shared theme until the passages themselves support that conclusion.
• Follow useful cross-references to build a biblical-theological framework while labeling the resulting synthesis as interpretation rather than dataset-attested relationship metadata.