@mistereman22/osrs-wiki-mcp-server 1.0.2 → 1.0.4

package/resources/page-structure.md

@@ -0,0 +1,66 @@
+# Page Structure Guide
+
+> Page types, key sections, infobox→bucket mapping for `get-page`.
+
+## Page Structure by Type
+
+| Page Type | Key Sections |
+|-----------|--------------|
+| **Items** | Infobox (stats), Obtaining, Usage, Changes |
+| **Monsters** | Infobox (stats), Drops, Strategies, Locations |
+| **Bosses** | Infobox, Mechanics, Strategies, Drops, Hard mode |
+| **Quests** | Infobox, Requirements, Walkthrough, Rewards, Quick guide |
+| **Locations** | Map, Features, NPCs, Monsters, Shops |
+| **Raids** | Overview, Entry requirements, Rooms/Bosses, Rewards, Strategies |
+
+## Infobox Fields (Semantic Data → Use Buckets)
+
+Most structured data lives in infoboxes and is available via buckets:
+
+| Page Type | Bucket | Key Fields |
+|-----------|--------|------------|
+| Items | `bucket("infobox_item")` | item_name, item_id, examine, high_alchemy_value, value, weight, buy_limit, is_members_only, tradeable, image, release_date |
+| Items (GE) | `bucket("exchange")` | name, value, high_alch, low_alch, buy_volume, sell_volume |
+| Monsters | `bucket("infobox_monster")` | name, combat_level, hitpoints, max_hit, slayer_level, slayer_experience, slayer_category, assigned_by, attack/defence bonuses, attack_speed, poisonous, attribute, attack_style, respawn_time, elemental_weakness |
+| Shops | `bucket("infobox_shop")` | shop_name, location, owner, specialty, shop_version, is_members_only |
+| NPCs | `bucket("infobox_npc")` | id, combat_level, examine |
+| Quests | `bucket("quest")` | name, requirements, rewards, difficulty, length, series |
+| Spells | `bucket("infobox_spell")` | name, level, rune_cost, effect, description |
+
+**Rule**: If you need infobox data → use the bucket, not the page.
+
+## Redirects
+
+Many pages redirect (e.g., "Abby whip" → "Abyssal whip"). `get-page` follows redirects by default. If you need the redirect target without following, use `followRedirects=false`.
+
+## Disambiguation
+
+Common disambiguation patterns:
+- `Item Name (disambiguation)` — rare
+- `Item Name (item)` vs `Item Name (monster)` — e.g., "Dragon (item)" vs "Dragon (monster)"
+- Use `search-page` to find the correct variant
+
+## Large Page Strategy
+
+For pages with many sections:
+
+1. `get-page "Title" metadata=true` — get section outline
+2. Identify needed section number
+3. `get-page "Title" section=N` — fetch only that section
+
+Example:
+```bash
+get-page "Chest (Tombs of Amascut)" metadata=true
+# Returns sections: Lead, Rewards, Unique drops, Standard drops, Supply drops
+get-page "Chest (Tombs of Amascut)" section=3  # Standard drops
+```
+
+## Quick Reference: Common Prefixes
+
+| Looking for... | Try prefix search |
+|----------------|-------------------|
+| Weapon variants | `search-page-by-prefix "Dragon "` |
+| Armour sets | `search-page-by-prefix "Bandos "` |
+| Potions | `search-page-by-prefix "Prayer "` |
+| Bosses | `search-page-by-prefix "Phantom "` |
+| Quests | `search-page-by-prefix "Recipe for "` |

package/resources/search-guide.md

@@ -0,0 +1,190 @@
+# Search Guide
+
+> Tool selection, broad-first search, canonical page heuristic, section discovery for `search-page`, `search-page-by-prefix`, `get-page`.
+
+## Tool Selection Decision Tree (START HERE)
+
+```
+User asks a question
+│
+├─ Structured data? (stats, prices, drops, requirements, gear, varbits)
+│   → bucket-query (read context://bucket-query-guide)
+│
+├─ Canonical list? (all bosses, all quests, all items, all updates)
+│   → get-category-members (read context://namespace-category-guide)
+│
+├─ Custom namespace list? (game updates in Update:, calculators)
+│   → get-site-info + get-category-members (read context://namespace-category-guide)
+│
+├─ Exact page title known?
+│   → get-page directly
+│
+├─ Know start of title? (e.g., "Abyssal ", "Dragon ")
+│   → search-page-by-prefix "Prefix"
+│
+└─ General topic / keywords / unstructured content?
+    → search-page with BROAD query + high limit (50-100)
+      → Check results for canonical index page → get-page that page
+```
+
+## Tool Comparison
+
+| Tool | Use Case | Returns |
+|------|----------|---------|
+| `search-page` | General topic, keyword, phrase | Page titles + snippets (full-text) |
+| `search-page-by-prefix` | Known prefix, autocomplete | Page titles only (no snippets) |
+| `get-page` (metadata) | Get section outline of known page | Section headings + numbers |
+
+## Broad-First Search Strategy
+
+**Start with a broad query + high limit to find the canonical page, then narrow.**
+
+| Approach | Query | Limit | Why |
+|----------|-------|-------|-----|
+| ❌ Too specific | `"2026 game updates"` | 10 | Misses "Game updates" page |
+| ✅ Broad first | `"game updates"` | 50-100 | Finds canonical page immediately |
+| ✅ Broad first | `"updates"` | 50 | Even broader, catches "Game updates" |
+
+### Why this works
+- `search-page` max limit = **100** (default 10). One broad call with `limit=100` costs same as narrow call.
+- Canonical index pages (e.g., "Game updates", "Quest list") rank high for broad terms.
+- Snippets in broad results reveal structure → you pick the right page → one `get-page`.
+
+### Workflow
+```
+1. search-page "updates" limit=100
+   → #2: "Game updates" (snippet: "chronological list of all game updates")
+2. get-page "Game updates"          ← canonical page, has everything
+3. (if needed) get-page "Game updates" section=3  ← specific year
+```
+
+### Anti-pattern
+```
+search-page "2026 game updates" limit=10  → misses canonical page
+search-page "Easter 2026 update"          → too narrow, fragments context
+search-page "Cow Boss update"             → too narrow
+```
+
+## search-page Best Practices
+
+### Do:
+- Use specific keywords: `"hydra gear setup"` not `"hydra"`
+- Include context: `"prayer flicking technique"` not `"prayer"`
+- Use quotes for exact phrases: `"/roll"` or `"drop rate"`
+
+### Don't:
+- Use single generic words: `"item"`, `"boss"`, `"quest"`
+- Overuse - check bucket-index first for structured data
+
+### Result Interpretation:
+- Snippet shows match context - verify it's the right page
+- Page ID/size/timestamp help gauge relevance
+- Namespace 0 = main content (what you want)
+
+## search-page-by-prefix Best Practices
+
+### Do:
+- Use when you know the prefix: `"Abyssal "`, `"Dragon "`, `"Barbarian "`
+- Short prefixes return more results: `"Dra"` vs `"Dragon "`
+- Combine with namespace filter for categories: `namespace: 14`
+
+### Don't:
+- Use for general topics (use `search-page`)
+- Expect snippets - only titles returned
+
+## get-page for Section Discovery
+
+### Before reading a large page:
+```bash
+get-page "Chest (Tombs of Amascut)" metadata=true
+```
+
+### Returns:
+```json
+{
+  "sections": [
+    {"index": 0, "title": "Lead"},
+    {"index": 1, "title": "Rewards"},
+    {"index": 2, "title": "Unique drops"},
+    {"index": 3, "title": "Standard drops"},
+    {"index": 4, "title": "Supply drops"}
+  ]
+}
+```
+
+### Then fetch only what you need:
+```bash
+get-page "Chest (Tombs of Amascut)" section=3   # Standard drops
+get-page "Chest (Tombs of Amascut)" section=2   # Unique drops
+```
+
+### Section 0 = Lead (intro text)
+### Section N = Heading N in the article
+
+## Common Search Patterns
+
+| Goal | Tool | Query |
+|------|------|-------|
+| Item stats/price | bucket-query | `bucket("exchange").select(...).where("name","...")` |
+| Monster stats | bucket-query | `bucket("infobox_monster").select(...).where("name","...")` |
+| Where item drops from | bucket-query | `bucket("dropsline").select(...).where("item_name","...")` |
+| What boss drops | get-page + metadata | `get-page "Boss Name" metadata=true` → `section=N` |
+| Quest requirements | bucket-query | `bucket("quest").select(...).where("name","...")` |
+| Gear for activity | bucket-query | `bucket("recommended_equipment").select(...).where("activity","...")` |
+| Shop location/items | bucket-query | `bucket("infobox_shop").select(...).where("shop_name","...")` |
+| General mechanic/guide | search-page | `search-page "prayer flicking"` |
+| Browse category | get-page | `get-page "Category:Bosses"` |
+| Find exact title variant | search-page-by-prefix | `search-page-by-prefix "Tumeken"` |
+
+## Canonical Page Heuristic
+
+### When search returns an obvious index/list page → FETCH IT IMMEDIATELY
+
+If a search result's **title** and **snippet** together indicate it's the canonical aggregation page for your topic, stop searching and `get-page` it.
+
+| Signal | Example |
+|--------|---------|
+| Title = plural/category name | `"Game updates"`, `"Quest list"`, `"Item sets"` |
+| Snippet shows dated/structured entries | `"lists Easter 2026, Cow Boss, Getting Around Poll in chronological order"` |
+| Title matches your query intent exactly | Search `"2026 updates"` → result `"Game updates"` |
+
+### Anti-pattern to avoid
+```
+search-page "Game update 2026"
+  → result: "Game updates" (snippet: "lists Easter 2026, Cow Boss...")
+  → ❌ KEEP SEARCHING for "2026 update" "Easter 2026" etc.
+  → ✅ get-page "Game updates"  ← canonical page, has everything
+```
+
+### Why this works
+- Wiki index pages (Game updates, Quest list, Monster drop tables) are **curated, complete, and maintained**
+- Search snippets often reveal the exact structure (chronological, by tier, by category)
+- One `get-page` > 5+ `search-page` calls (rate limit, context, accuracy)
+
+### Quick test
+> "Does this result look like *the* page a human would bookmark for this topic?" → Yes → `get-page`
+
+## Rate Limit Awareness
+
+- Each search tool call counts toward MediaWiki rate limit (30/min)
+- Prefer bucket-query for structured data (same limit, but returns exact data)
+- Use `metadata=true` + `section=N` to minimize page read calls
+- Cache results mentally within a conversation - don't re-search
+
+## DPL / Template Pages
+
+If `get-page` returns `<dpl>` tags or `{{Template}}` transclusions — **the data isn't there**. See `context://namespace-category-guide` → "Dynamic Content Patterns" for the workflow:
+
+1. `get-site-info` → find namespace ID
+2. `get-category-members` with category + namespace
+3. `get-pages` for details
+
+Common on OSRS: Game updates (`Update:` namespace 1198), calculators, module outputs.
+
+## Fallback Chain
+
+1. **bucket-query** (if structured data exists) → check `context://bucket-catalog`
+2. **search-page-by-prefix** (if you know title prefix)
+3. **search-page** (general topic)
+4. **get-page + metadata** (large page, need specific section)
+5. **get-page** (full page, last resort)

package/resources/tool-selection.md

@@ -0,0 +1,99 @@
+# Tool Selection Guide (READ FIRST)
+
+> Decide which tool to use, then read the corresponding guide.
+
+## Decision Tree
+
+```
+User asks a question
+│
+├─ Structured data? (stats, prices, drops, requirements, gear, varbits, IDs)
+│   → bucket-query
+│   → Read: context://bucket-query-guide
+│
+├─ Canonical list? (all bosses, all quests, all items, all updates)
+│   → get-category-members
+│   → Read: context://namespace-category-guide
+│
+├─ Custom namespace list? (game updates, calculators)
+│   → get-site-info + get-category-members
+│   → Read: context://namespace-category-guide
+│
+├─ Exact title known?
+│   → get-page
+│
+├─ Know title prefix? (e.g., "Abyssal ", "Dragon ")
+│   → search-page-by-prefix
+│   → Read: context://search-guide
+│
+└─ General topic / unstructured content? (lore, mechanics, guides)
+    → search-page (broad query + high limit)
+    → Check for canonical page → get-page
+    → Read: context://search-guide
+```
+
+## Quick Tool → Resource Map
+
+| Tool | Primary Resource | Secondary |
+|------|------------------|-----------|
+| `bucket-query` | `context://bucket-query-guide` | `context://bucket-catalog` |
+| `get-category-members` | `context://namespace-category-guide` | — |
+| `get-site-info` | `context://namespace-category-guide` | — |
+| `get-page` | `context://search-guide` (section discovery) | `context://page-structure` |
+| `search-page` | `context://search-guide` | — |
+| `search-page-by-prefix` | `context://search-guide` | — |
+| `get-links-here` | `context://namespace-category-guide` | — |
+
+## Most Common Patterns (Copy-Paste)
+
+### Bucket Query (structured data)
+```lua
+-- Item price/stats
+bucket("exchange").select("name","value","high_alch").where("name","Abyssal whip").run()
+bucket("infobox_item").select("item_name","value","high_alch","weight").where("item_name","Abyssal whip").run()
+
+-- Monster stats
+bucket("infobox_monster").select("name","combat_level","hitpoints","max_hit").where("name","Hydra").run()
+
+-- Drop sources for item
+bucket("dropsline").select("item_name","drop_json").where("item_name","Abyssal whip").run()
+
+-- Quest requirements
+bucket("quest").select("name","requirements").where("name","Dragon Slayer I").run()
+
+-- Gear for activity
+bucket("recommended_equipment").select("item","slot").where("activity","hydra").run()
+
+-- Varbit
+bucket("varbit").select("name","value").where("name","hydra_task").run()
+
+-- Shop
+bucket("infobox_shop").select("shop_name","location","owner").where("shop_name","Bob's Brilliant Axes").run()
+```
+
+### Category Members (canonical lists)
+```bash
+# All bosses
+get-category-members "Bosses"
+
+# All quests
+get-category-members "Quests"
+
+# All game updates (Update: namespace 1198)
+get-category-members "Updates" namespaces=[1198]
+
+# All GE items
+get-category-members "Grand Exchange items"
+```
+
+### Search (unstructured content)
+```bash
+# Broad first → canonical page → section
+search-page "updates" limit=100
+get-page "Game updates"
+get-page "Game updates" section=3
+
+# Section discovery
+get-page "Chest (Tombs of Amascut)" metadata=true
+get-page "Chest (Tombs of Amascut)" section=3
+```

package/resources/bucket-index.md

@@ -1,147 +0,0 @@
-# Bucket Index (hint list)
-
-> On-demand hints for `bucket-query` on the Old School RuneScape Wiki.
-> **This index is not authoritative.** Purpose descriptions are best-effort
-> summaries; fields always come from the schema page. If nothing here fits
-> your need, or a name fails at runtime, fall back to discovery:
-> `search-page-by-prefix` in the `Bucket:` namespace (id 9592), then
-> `get-page` on `Bucket:<Name>` to see exact fields and a data preview.
->
-> Naming rule: pass the page title lowercased with spaces as underscores
-> (`Bucket:Clue equipment` -> `bucket("clue_equipment")`). All bucket and
-> field names are lowercase; fields are verified on the schema page, never
-> guessed from this file.
-
-## Known buckets
-
-| Query as | Page | Holds |
-| --- | --- | --- |
-| `bucket("bountytaskline")` | Bucket:Bountytaskline | Bounty Hunter task line entries |
-| `bucket("clue_equipment")` | Bucket:Clue equipment | Equipment associated with clue scroll steps |
-| `bucket("clue_info")` | Bucket:Clue info | Clue scroll metadata/step info |
-| `bucket("collection_log_source")` | Bucket:Collection log source | Collection log sources per item: item_id, item_name, sources (repeated wikilinks), rates, kinds. **Only queryable by item_name** - cannot filter by source. To find "what drops from X", use the source page directly. |
-| `bucket("combat_achievement")` | Bucket:Combat achievement | Combat achievement tasks, tiers, requirements |
-| `bucket("couriertaskline")` | Bucket:Couriertaskline | Courier activity task line entries |
-| `bucket("demonicpactleaguetask")` | Bucket:Demonicpactleaguetask | Demonic Pact league task entries |
-| `bucket("dependency_list")` | Bucket:Dependency list | Item/object dependency relations |
-| `bucket("drop_table_sources")` | Bucket:Drop table sources | Shared drop tables (rare drop table, gem drop table, herb drop table) and their roll/rarity data. Not specific monster/boss drop tables. |
-| `bucket("dropsline")` | Bucket:Dropsline | Drop table line entries: item_name, drop_json (JSON text with "Dropped from", rarity, quantity), rare_drop_table. **Only queryable by item_name** - cannot filter by source. To find "what drops from X", use the source page (boss/chest/activity) directly. |
-| `bucket("exchange")` | Bucket:Exchange | Grand Exchange prices (name, value, high/low, volumes) |
-| `bucket("feedback")` | Bucket:Feedback | Wiki feedback entries (site-internal) |
-| `bucket("ge_index_header")` | Bucket:Ge index header | Grand Exchange index header/snapshot metadata |
-| `bucket("infobox_activity")` | Bucket:Infobox activity | Skilling/activity infobox data |
-| `bucket("infobox_bonuses")` | Bucket:Infobox bonuses | Equipment combat stats and bonuses |
-| `bucket("infobox_construction")` | Bucket:Infobox construction | Construction object/buildable data |
-| `bucket("infobox_grid_master_unlock")` | Bucket:Infobox grid master unlock | Grid Master unlock data |
-| `bucket("infobox_item")` | Bucket:Infobox item | Core item data: item_name, item_id, examine, high_alchemy_value, value, weight, buy_limit, is_members_only, tradeable, image, league_region, release_date, removal_date, value, weight, high_alchemy_value, buy_limit, default_version, quest, version_anchor, is_members_only, tradeable |
-| `bucket("infobox_location")` | Bucket:Infobox location | Location/area data |
-| `bucket("infobox_monster")` | Bucket:Infobox monster | Monster combat stats: name, combat_level, hitpoints, max_hit, slayer_level, slayer_experience, slayer_category, assigned_by, attack/defence bonuses (stab/slash/crush/magic/range), attack_speed, hitpoints, max_hit, poisonous, attribute, attack_style, attack_speed, respawn_time, experience_bonus, flat_armour, magic_resistance, size, attack/strength/defence/ranged/magic levels, attack_bonus, strength_bonus, range_strength_bonus, magic_damage_bonus, poison/venom/freeze/burn/cannon/thrall resistance, elemental_weakness, elemental_weakness_percent |
-| `bucket("infobox_npc")` | Bucket:Infobox npc | NPC data: id, combat level, examine |
-| `bucket("infobox_pure")` | Bucket:Infobox pure | Pure-entity infobox data (verify schema) |
-| `bucket("infobox_scenery")` | Bucket:Infobox scenery | Scenery/object entity data |
-| `bucket("infobox_ship_part")` | Bucket:Infobox ship part | Ship part data |
-| `bucket("infobox_shop")` | Bucket:Infobox shop | Shop data: shop_name, location, owner, specialty, shop_version, is_members_only |
-| `bucket("infobox_spell")` | Bucket:Infobox spell | Spell data: requirements, effects, costs |
-| `bucket("interface")` | Bucket:Interface | Game interface definitions |
-| `bucket("item_id")` | Bucket:Item id | Item name to id mappings |
-| `bucket("locline")` | Bucket:Locline | Location line entries |
-| `bucket("logs")` | Bucket:Logs | Woodcutting log/tree data |
-| `bucket("map")` | Bucket:Map | Map marker/region data |
-| `bucket("mine")` | Bucket:Mine | Mining node/rock data |
-| `bucket("mmg_output")` | Bucket:Mmg output | Money making guide output calculations |
-| `bucket("money_making_guide")` | Bucket:Money making guide | Money making methods: profit, requirements |
-| `bucket("music")` | Bucket:Music | Music track unlock data |
-| `bucket("music_file")` | Bucket:Music file | Music file metadata |
-| `bucket("music_map")` | Bucket:Music map | Music tracks by region/map area |
-| `bucket("music_versions")` | Bucket:Music versions | Music version/history entries |
-| `bucket("npc_id")` | Bucket:Npc id | NPC name to id mappings |
-| `bucket("object_id")` | Bucket:Object id | Object/scenery id mappings |
-| `bucket("quest")` | Bucket:Quest | Quest data: requirements, rewards, difficulty |
-| `bucket("recipe")` | Bucket:Recipe | Creation/processing recipes: inputs, outputs, skills |
-| `bucket("recommended_equipment")` | Bucket:Recommended equipment | Recommended gear setups by activity/slot (schema not available on page - use runtime discovery) |
-| `bucket("seachart")` | Bucket:Seachart | Sea chart/naval data |
-| `bucket("sound_effect")` | Bucket:Sound effect | Sound effect metadata |
-| `bucket("soundtrack")` | Bucket:Soundtrack | Soundtrack track data |
-| `bucket("storeline")` | Bucket:Storeline | Store line entries |
-| `bucket("testing_bucket")` | Bucket:Testing bucket | Wiki-internal testing data (not real content) |
-| `bucket("transcript")` | Bucket:Transcript | NPC/dialogue transcripts |
-| `bucket("update")` | Bucket:Update | Game update records: dates, summaries |
-| `bucket("varbit")` | Bucket:Varbit | Varbit id to meaning mappings |
-
-## Query reminders
-
-- Query must end in `.run()`; `limit` max 500, paginate with `continueFrom`.
-- Equality: `where("field","value")`; operators: `where({"field",">",value})`
-  with `=`, `!=`, `>`, `<`, `>=`, `<=`; compose with `Bucket.And/Or/Not`.
-- Example: `bucket("infobox_item").select("item_id","examine").where("item_name","Raw lobster").run()`
-- Full Lua reference: https://meta.weirdgloop.org/w/Extension:Bucket/Usage
-
-## Best Practices (MANDATORY)
-
-- **ALWAYS use `.select("field1","field2")`** - only fetch needed columns
-- **ALWAYS use `.where("field","value")`** - filter rows, never fetch all
-- **Use `.limit(N)`** - cap results (max 500)
-- **Chain filters** with `Bucket.And/Or/Not` for complex conditions
-- **NEVER** run `.run()` without `.select()` and `.where()`
-
-### Exact Match Only - No Wildcards
-
-bucket-query uses **exact match only**. To find exact values:
-1. `search-page` for the term → get exact name
-2. Use exact name in bucket query
-
-Example: "Bob's Axe Shop" → `search-page "bob axe"` → "Bob's Brilliant Axes" → `bucket("infobox_shop").where("shop_name","Bob's Brilliant Axes")`
-
-### Example Patterns
-
-✅ Good:
-```
-bucket("exchange").select("name","value","high_alch").where("name","Abyssal whip").run()
-bucket("infobox_shop").select("shop_name","location").where("location","Lumbridge").limit(5).run()
-```
-
-❌ Bad (causes context bloat):
-```
-bucket("infobox_shop").run()
-bucket("exchange").select("*").run()
-```
-
-### Pagination
-```lua
-bucket("infobox_shop").select("shop_name","location").limit(50).continueFrom("cursor").run()
-```
-
-### Common Bucket Patterns (Always use .select() and .where())
-
-| Task | Bucket Query |
-|------|--------------|
-| GE prices | `bucket("exchange").select("name","value","high_alch").where("name","Abyssal whip").run()` |
-| Monster stats | `bucket("infobox_monster").select("name","combat_level","hitpoints").where("name","Hydra").run()` |
-| Item stats | `bucket("infobox_item").select("item_name","value","high_alch").where("item_name","Abyssal whip").run()` |
-| Drop sources | `bucket("dropsline").select("source","quantity").where("item","Abyssal whip").run()` |
-| Quest requirements | `bucket("quest").select("name","requirements").where("name","Dragon Slayer I").run()` |
-| Varbits | `bucket("varbit").select("name","value").where("name","hydra_task").run()` |
-| Shop lookup | `bucket("infobox_shop").select("shop_name","location","owner").where("shop_name","Bob's Brilliant Axes").run()` |
-| Gear recommendations | `bucket("recommended_equipment").select("item","slot").where("activity","hydra").run()` |
-
-### Workflow: Find Exact Values First
-
-1. **search-page** for term → get exact name
-2. Use exact name in `bucket-query` with `.where()`
-
-Example: "best gear for hydra" → `bucket("recommended_equipment").select("item","slot").where("activity","hydra").run()`
-
-### Raid/Group Drop Patterns (use get-page, NOT buckets)
-
-No bucket supports "what drops from source" queries. For raid/group content, use `get-page` on individual source pages:
-
-| Content | Query |
-|---------|-------|
-| Tombs of Amascut (ToA) | `get-page "Chest (Tombs of Amascut)"` then `get-page "Tumeken's Warden"`, `get-page "Akkha"`, etc. |
-| Chambers of Xeric (CoX) | `get-page "Chambers of Xeric"` then individual boss pages (Olm, Great Olm, etc.) |
-| Theatre of Blood (ToB) | `get-page "Theatre of Blood"` then individual boss pages (Maiden, Bloat, Nylocas, Sotetseg, Xarpus, Verzik) |
-| General boss drops | `get-page "<Boss Name>"` (e.g., "Zulrah", "Vorkath", "Phantom Muspah") |
-
-For large pages: `get-page "<Title>" metadata=true` → get section outline → `get-page "<Title>" section=N` for drops section.
-
-Use page reads (`get-page`, `search-page`) ONLY for unstructured content: lore, mechanics, walkthroughs.

package/resources/search-strategies.md

@@ -1,111 +0,0 @@
-# Search Strategies
-
-> When and how to use each search tool effectively.
-
-## Tool Comparison
-
-| Tool | Use Case | Returns |
-|------|----------|---------|
-| `search-page` | General topic, keyword, phrase | Page titles + snippets (full-text) |
-| `search-page-by-prefix` | Known prefix, autocomplete | Page titles only (no snippets) |
-| `get-page` (metadata) | Get section outline of known page | Section headings + numbers |
-
-## Decision Tree
-
-```
-Need to find a page?
-│
-├─ Know exact title? → get-page "Exact Title"
-│
-├─ Know start of title? → search-page-by-prefix "Prefix"
-│
-├─ General topic / keywords? → search-page "keywords"
-│
-└─ Want to browse a category? → get-page "Category:Name"
-```
-
-## search-page Best Practices
-
-### Do:
-- Use specific keywords: `"hydra gear setup"` not `"hydra"`
-- Include context: `"prayer flicking technique"` not `"prayer"`
-- Use quotes for exact phrases: `"/roll"` or `"drop rate"`
-
-### Don't:
-- Use single generic words: `"item"`, `"boss"`, `"quest"`
-- Overuse - check bucket-index first for structured data
-
-### Result Interpretation:
-- Snippet shows match context - verify it's the right page
-- Page ID/size/timestamp help gauge relevance
-- Namespace 0 = main content (what you want)
-
-## search-page-by-prefix Best Practices
-
-### Do:
-- Use when you know the prefix: `"Abyssal "`, `"Dragon "`, `"Barbarian "`
-- Short prefixes return more results: `"Dra"` vs `"Dragon "`
-- Combine with namespace filter for categories: `namespace: 14`
-
-### Don't:
-- Use for general topics (use `search-page`)
-- Expect snippets - only titles returned
-
-## get-page for Section Discovery
-
-### Before reading a large page:
-```bash
-get-page "Chest (Tombs of Amascut)" metadata=true
-```
-
-### Returns:
-```json
-{
-  "sections": [
-    {"index": 0, "title": "Lead"},
-    {"index": 1, "title": "Rewards"},
-    {"index": 2, "title": "Unique drops"},
-    {"index": 3, "title": "Standard drops"},
-    {"index": 4, "title": "Supply drops"}
-  ]
-}
-```
-
-### Then fetch only what you need:
-```bash
-get-page "Chest (Tombs of Amascut)" section=3   # Standard drops
-get-page "Chest (Tombs of Amascut)" section=2   # Unique drops
-```
-
-### Section 0 = Lead (intro text)
-### Section N = Heading N in the article
-
-## Common Search Patterns
-
-| Goal | Tool | Query |
-|------|------|-------|
-| Item stats/price | bucket-query | `bucket("exchange").select(...).where("name","...")` |
-| Monster stats | bucket-query | `bucket("infobox_monster").select(...).where("name","...")` |
-| Where item drops from | bucket-query | `bucket("dropsline").select(...).where("item_name","...")` |
-| What boss drops | get-page + metadata | `get-page "Boss Name" metadata=true` → `section=N` |
-| Quest requirements | bucket-query | `bucket("quest").select(...).where("name","...")` |
-| Gear for activity | bucket-query | `bucket("recommended_equipment").select(...).where("activity","...")` |
-| Shop location/items | bucket-query | `bucket("infobox_shop").select(...).where("shop_name","...")` |
-| General mechanic/guide | search-page | `search-page "prayer flicking"` |
-| Browse category | get-page | `get-page "Category:Bosses"` |
-| Find exact title variant | search-page-by-prefix | `search-page-by-prefix "Tumeken"` |
-
-## Rate Limit Awareness
-
-- Each search tool call counts toward MediaWiki rate limit (30/min)
-- Prefer bucket-query for structured data (same limit, but returns exact data)
-- Use `metadata=true` + `section=N` to minimize page read calls
-- Cache results mentally within a conversation - don't re-search
-
-## Fallback Chain
-
-1. **bucket-query** (if structured data exists) → check `context://bucket-index`
-2. **search-page-by-prefix** (if you know title prefix)
-3. **search-page** (general topic)
-4. **get-page + metadata** (large page, need specific section)
-5. **get-page** (full page, last resort)