region-research

name	region-research
description	Research Elder Scrolls lore for a region and propose adventure themes with matching card catalog decks. Use when planning new adventures for a region — e.g. "research Morrowind adventures", "what adventures could we do in Cyrodiil". Takes a region name as argument.
model	opus
effort	high

Region Adventure Research

Research Elder Scrolls lore for a specific region and propose Path of Tamriel adventure themes, mapping lore to the existing card catalog to produce actionable adventure designs.

Arguments

$ARGUMENTS — A region name (e.g., Hammerfell, Morrowind, Cyrodiil). May optionally include a wiki URL or specific sub-region focus.

Phase 1: Gather Lore

1a. Check existing research

Look in development-artifacts/ for any existing research files for this region (e.g., {region}_adventure_research.md). If one exists, read it — you may be updating rather than starting fresh.

1b. Check the region config and map covered storylines

Read data/regions/{region_id}.json to see the current state — what adventures already exist, the region's display name, and map image path.

If the region already has adventures, read each adventure JSON and its associated player deck to identify which lore storylines are already covered. For example, an adventure with a Companions deck and Jorrvaskr references = the Companions questline is covered. Map each existing adventure to its lore source so Phase 1c research can focus exclusively on uncovered storylines. This avoids wasting time researching ground already covered.

1c. Research lore from web sources

WebSearch snippets are usually sufficient. Multiple targeted web searches (one for geography, one for history, one for cultural traditions, one for ESO storylines, etc.) will typically surface enough lore from result summaries alone — often more efficiently than fetching full wiki pages. Launch 5-8 parallel web searches as the primary lore-gathering strategy.

Only use WebFetch if a specific page looks exceptionally promising from search results. Most wiki sites (UESP, Fandom, Imperial Library) frequently return 403 errors on direct fetch, so don't spend time on fetch-then-fallback chains. Web searches are the workhorse here.

Parallelise aggressively — launch all initial web searches concurrently, not sequentially.

Gather information across these categories:

Geography & Settings

• Major cities and their political/cultural significance
• Distinct sub-regions (deserts, forests, mountains, coasts, ruins)
• Notable landmarks, dungeons, ruins (especially Dwemer, Ayleid, Daedric)

History & Storylines

• Major wars, conflicts, and political events
• Storylines from TES games set in this region (main quests, guild questlines, DLC)
• Era-spanning narratives (which events span First through Fourth Era?)
• Faction conflicts and civil wars

Prominent Figures

• Legendary heroes, villains, kings, generals, mages
• Characters from TES game storylines set here
• Daedric Princes with strong regional ties
• Guild leaders, faction heads, historical warriors

Races & Factions

• Dominant race(s) and their culture
• Internal factions and political divisions
• Enemy factions (invaders, cults, undead, Daedra)
• Guilds and organizations based in the region

Culture & Themes

• Unique cultural practices (e.g., Sword-Singing for Hammerfell, Tribunal worship for Morrowind)
• Religious traditions and pantheon
• Architectural and aesthetic identity
• Signature combat styles or magic traditions

1d. Research sub-topics in depth

For the most promising storylines, do additional targeted searches. For example:

• If the region has a unique warrior tradition, research its history, notable practitioners, and ranks
• If there's a famous war, research key battles and commanders
• If there's a prominent game storyline, research its full plot arc and major characters
• ESO zone storylines — If the region has ESO zones (most do), search for each zone's main storyline. ESO zones often have self-contained narrative arcs with named villains, faction conflicts, and set-piece battles that translate directly into adventure proposals. Search for "Elder Scrolls Online {zone_name} storyline" for each zone in the region.

Phase 2: Catalog Audit

2a. Find region-themed cards

Search src/deck/card_catalog.gd for cards connected to the region:

1. By race subtype — grep for the dominant race (e.g., "Redguard", "Dark Elf", "Nord")
2. By place names — grep for city/region names that appear in card names (e.g., Alik'r, Sentinel, Rihad, Craglorn)
3. By cultural keywords — grep for faction/culture terms (e.g., Yokud, Crown, Ansei, Sword)
4. By lore figure names — grep for named characters discovered in Phase 1
5. By regional creature types — grep for creature subtypes native to the region (e.g., "Reptile", "Wamasu", "Lurcher" for Black Marsh; "Dragon", "Draugr" for Skyrim). These are essential for building thematic enemy decks

Note: Grep results on card_catalog.gd often produce [Omitted long matching line] for seed entries. When this happens, use Read with the specific line offset to see the full card data. Also check data/wiki_scrape/cards.json — it contains pre-scraped card data that may be easier to search for lore cross-references.

2b. Identify legendary cards

From the matches, extract all legendary, unique cards with the region's dominant race subtype. These are potential deck anchors or boss candidates.

2c. Map card themes

Group the discovered cards into thematic clusters:

• Item/equipment synergy cards
• Rally/army-building cards
• Magic/spell synergy cards
• Tribal synergy cards (race-specific)
• Keyword-themed groups (Slay, Pilfer, Ward, etc.)

Note which clusters have enough depth (8+ cards) to support a full adventure deck. A viable deck needs ~30 cards with a coherent mechanical identity.

2d. Check for named lore cards

Look for cards that directly reference lore events or artifacts (e.g., "Siege of Stros M'Kai", "The Red Year"). These provide strong thematic anchors.

2e. Audit enemy card pools

Search the catalog for cards that could populate enemy decks — this is just as important as finding player cards. Common enemy pools:

• Daedra subtype (for Daedric invasion themes)
• Skeleton/Spirit/Vampire subtypes (for undead/necromancy themes)
• Beast/Animal subtypes native to the region
• Ward/Mage cards in Intelligence (for enemy mage decks)
• Any race subtype that could serve as an enemy faction

Note the card count per enemy theme. A viable enemy deck needs ~12-15 distinct cards (with quantity multipliers reaching ~30 total). If an enemy theme has fewer than 8 cards, flag it as insufficient.

Tip: Use a subagent (Agent tool with subagent_type: Explore) for the full catalog audit. Searching card_catalog.gd and data/wiki_scrape/cards.json across 10+ grep patterns is a perfect subagent task — it keeps the main context clean and runs faster via parallel searches.

2f. Review existing enemy decks

Read enemy deck files in data/decks/adventure/enemies/ to understand:

• What enemy themes are already built (e.g., Thalmor, Necromancer, Draugr)
• Enemy deck structure: card count, quantity distribution, attribute spread
• Which cards are already "claimed" by existing enemy decks (avoid reusing the same deck wholesale)

This prevents proposing an adventure with enemy decks that duplicate what's already in the game.

Also check player decks — note which legendary/unique cards are already included in existing adventure player decks. A card used as a deck anchor in one adventure should not be proposed as a boss or anchor for another adventure in the same region (e.g., Miraak in the Dragon deck shouldn't also be the boss of a Dragonborn DLC adventure).

Phase 3: Cross-Reference

Run Phase 1 (lore) and Phase 2 (catalog) in parallel — catalog greps don't depend on lore research completing first. Start grepping for the region's dominant race and obvious place names while lore searches are still in flight. Phase 3 cross-referencing begins once both are done.

3a. Match figures to cards

For each prominent lore figure identified in Phase 1, check if they exist as a card in the catalog. Note:

• Exists as legendary card → natural boss or deck anchor
• Exists as non-legendary card → can still feature in the adventure narrative
• Does not exist → potential custom boss (enemy-only deck), note for future card imports

3b. Match storylines to card pools

For each promising storyline, assess whether the card catalog has enough thematically matching cards to build:

• A player deck (30 cards, coherent theme, 1-3 attributes — tri-color is valid if the region has a tri-color faction like the Aldmeri Dominion)
• Enemy decks for 3-5 combat encounters (can reuse cards across enemies)
• At least one boss with a signature card or mechanic

3c. Review existing adventures and decks for patterns

Read 1-2 existing adventure JSON files in data/adventures/ to understand:

• Node count and type distribution (how many combats, events, shops, etc.)
• Enemy deck naming conventions
• XP reward ranges
• Boss health ranges and quality values

Critical: Read ALL existing player deck files in data/decks/adventure/ (excluding enemies/) across all regions, not just the current one. Note the attribute_ids of each. Map out which attribute pairs are already used globally. Every adventure proposal should target an unused attribute pair (or a tri-color combo) to ensure mechanical variety across the entire adventure catalog.

Phase 4: Propose Adventures

For each proposed adventure, provide:

Adventure Header

• Name — evocative title referencing the storyline
• Setting — specific locations within the region, sequenced as a journey
• Storyline — 2-3 sentence narrative arc

Deck Design

• Archetype — attribute combination and mechanical identity (e.g., "STR/INT Battlemage — item synergy")
• Core cards — list 12-15 specific cards from the catalog by name, explaining why each fits
• Legendary anchor — which legendary card(s) define the deck
• Mechanical coherence — explain why these cards work together

Boss & Enemies

• Final boss — name, lore justification, potential special mechanic
• Mini-boss — name, role in the storyline
• Enemy deck themes — what enemy decks represent (e.g., "Imperial soldiers", "Dwemer automatons")
• Enemy races — what subtypes enemies should use

Adventure Nodes

• Boons — themed boon names with flavor
• Events — 1-2 narrative choice events with mechanical consequences
• Shop/Reinforcement — themed location names
• Augment nodes — what augment flavor fits (e.g., "Dwemer Forge", "Enchanter's Altar")

Feasibility Rating

Rate each adventure on:

• Card pool depth (Excellent / Good / Moderate / Thin) — enough cards for a coherent deck?
• Boss availability (Ready / Needs import / Custom only) — is there a legendary card for the boss?
• Lore richness (Deep / Solid / Surface) — how much narrative material is there?
• Mechanical uniqueness (High / Moderate / Low) — does this play differently from existing adventures?
• Suggested difficulty (1-3) — matches the difficulty field in adventure JSON

Phase 5: Write Report

5a. Write the research document

Save the complete findings to development-artifacts/{region_id}_adventure_research.md with these sections:

1. Lore Overview — geography, history, culture summary
2. Key Regions & Cities — table of locations with notes
3. Races & Peoples — who lives here
4. Key Factions — political and military groups
5. Prominent Figures — table with name, role, and boss potential
6. Signature Cultural Elements — what makes this region unique
7. Cards in the Catalog — organized by legendary cards, place-named cards, and thematic groups
8. Existing Adventures & Attribute Usage — table of all current adventure decks with their attribute pairs, plus list of unused pairs. This is critical context for ensuring proposals don't overlap.
9. Adventure Proposals — 3-5 proposals with full details from Phase 4
10. Summary Table — ranked comparison of all proposals

5b. Prioritize and recommend

End with a clear recommendation for which adventure to build first, based on:

1. Card catalog depth (can we build a deck today?)
2. Lore iconicness (will players recognize and enjoy this storyline?)
3. Boss quality (is there a satisfying climactic fight?)
4. Mechanical uniqueness (does this play differently from existing adventures?)

Tips & Pitfalls

• UESP may block requests — always have fallback sources. Web searches often surface enough lore even when wikis are inaccessible.
• Card catalog is the constraint — exciting lore means nothing if the catalog can't support a deck. Always verify card pool depth before proposing an adventure.
• Check existing adventures for overlap — don't propose a theme too similar to an existing adventure in another region. Read data/adventures/*.json headlines. Also read existing player deck files in data/decks/adventure/ to check attribute overlap — if an existing deck already uses INT/WIL, differentiate your proposal with different attributes, a tri-color build, or a distinctly different mechanical identity (e.g., token/buff vs action-spam).
• Dual-attribute cards matter — a STR/INT deck can use STR cards, INT cards, STR/INT dual cards, and NEU cards. It cannot use STR/WIL or INT/AGI duals — both attributes must match the deck's pair. Don't miss the dual-attribute section of the catalog, but don't accidentally include off-pair duals in proposals.
• Race subtypes are key — the dominant race's subtype (e.g., "Redguard", "Dark Elf") is the strongest filter for finding thematically appropriate cards.
• Items, keywords, and effects cluster by attribute — STR has items/weapons, INT has wards/spells, AGI has movement/pilfer, END has guards/last gasp, WIL has tokens/healing. Match the region's cultural flavor to the right attribute.
• Boss fights need a hook — the best bosses have a unique mechanic or gimmick that reflects their lore identity, not just high stats.