This read-only OSRS Wiki MCP server lets AI clients research Old School RuneScape Wiki content with source-backed, bounded structured responses.
Search the Wiki for canonical pages and snippets, with pagination.
Read cleaned Wiki pages and list or read individual sections by numeric index.
Look up item facts such as description, properties, bonuses, alchemy values, and creation details.
Find item acquisition data: complete shop listings, monster drop sources, and a bounded overview of drops, shops, recipes, and ground spawns.
Get quest requirements without evaluating a player's progress.
Get separate monster variants, stats, map points, and access notes without combining variants or calculating DPS.
Receive provenance on successful responses, including Wiki URLs, fetch timestamps, attribution/license details, and exact revision links for parsed pages.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@osrs-wiki-mcp find drop sources for abyssal whip"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
A local, read-only Model Context Protocol server for the Old School RuneScape Wiki. It gives AI clients ten focused tools for Wiki research, with validated structured output and source provenance on every successful response.
Stateless: no player account, database, telemetry, or disk cache.
Source-backed: results include canonical Wiki URLs and fetch timestamps; parsed pages also include an exact revision link.
Bounded: response sizes, pagination, retries, and request time are capped.
Read-only: the server only retrieves public Wiki data.
Node.js 24 or newer and npx are required.
codex plugin marketplace add SSanderV/osrs-wiki-mcp --ref v1.1.4
codex plugin add osrs-wiki-mcp@osrs-wikiclaude plugin marketplace add SSanderV/osrs-wiki-mcp@v1.1.4 --scope user
claude plugin install osrs-wiki-mcp@osrs-wiki --scope usergemini extensions install https://github.com/SSanderV/osrs-wiki-mcp --ref v1.1.4The plugin and extension include the MCP setup plus a small Wiki-research
skill. All three launch the pinned npm package osrs-wiki-mcp@1.1.4.
Configure a standard-input/output server with:
{
"command": "npx",
"args": ["--yes", "osrs-wiki-mcp@1.1.4"]
}Start a fresh client session after installation. If you previously configured
an osrs-wiki MCP server directly, remove that registration first so the
plugin is the only provider of the server.
Related MCP server: OSRS MCP Server
Tool | Inputs | Use it for |
|
| Find canonical pages and snippets. |
|
| Read a bounded, cleaned page with its section index. |
|
| List a page's sections. |
|
| Read one section by its numeric index. |
|
| Get an item's description, properties, bonuses, and creation facts. |
|
| Page through complete shop listings. |
|
| Page through complete monster-drop listings. |
|
| Get a bounded overview of drops, shops, recipes, and ground spawns. |
|
| Get Wiki-sourced requirements without evaluating a player. |
|
| Get separate monster variants, map points, and access notes. |
Example requests:
“What are the requirements for Dragon Slayer I?”
“Where can an ironman obtain a rune scimitar?”
“Find the exact Wiki section explaining Hespori mechanics.”
All public text inputs accept 1–256 Unicode characters. Tool schemas declare
the exact limits and defaults. When results are truncated or upstream
pagination is incomplete, warnings explains the safe follow-up action.
Successful calls return:
content for clients that display readable text; and
structuredContent, validated against the tool's MCP output schema.
Provenance includes the contributing Wiki URLs, original fetchedAt time,
attribution, license details, and whether content was transformed. Parsed-page
responses additionally include the exact revision ID and revision URL.
Failures use stable error codes such as NOT_FOUND, UPSTREAM_TIMEOUT,
UPSTREAM_RATE_LIMITED, UPSTREAM_UNAVAILABLE,
UPSTREAM_INVALID_RESPONSE, and RESPONSE_TOO_LARGE. Invalid arguments remain
standard JSON-RPC parameter errors.
A tool call has a 30-second total budget; each upstream request gets at most 20 seconds within that budget.
Transient network and rate-limit failures receive at most two bounded
retries, respecting Retry-After only when time remains.
Responses are capped at 5 MiB. Large Bucket scans stop after 10,000 raw rows and report incomplete or truncated results.
Valid upstream responses may be cached in memory for five minutes. The cache is limited per process, never written to disk, and cleared on restart.
Standard output is reserved for MCP traffic. Sanitized diagnostics use standard error.
The server deliberately does not provide player progress, hiscores, Grand Exchange prices, DPS calculations, Wiki images, persistent storage, hosting, or a local Wiki mirror. Quest requirements are facts rather than met/missing evaluations, and monster variants are never combined.
npm.cmd ci --ignore-scripts
npm.cmd run typecheck
npm.cmd test
npm.cmd run build
npm.cmd run smoke:stdio
npm.cmd run pack:checkTests use synthetic fixtures and do not contact the live Wiki. The optional live smoke makes two low-volume requests and does not persist responses:
npm.cmd run test:liveSee CONTRIBUTING.md for development guidance and SECURITY.md for private vulnerability reporting.
The source code is MIT licensed. Retrieved Old School RuneScape Wiki content remains subject to CC BY-NC-SA 3.0, including its non-commercial and ShareAlike conditions. Downstream users are responsible for how they reuse that content. See THIRD_PARTY_NOTICES.md.
RuneScape and Old School RuneScape are trademarks of Jagex Limited. This independent project is not affiliated with, endorsed by, or sponsored by Jagex Limited, Weird Gloop, or the Old School RuneScape Wiki.
Get the complete paginated Old School RuneScape Wiki drop listing for an item; get_item_sources is the bounded overview.
Inputs: item (required), limit, offset. Returns: item, limit, total, offset, +11 more
Get the complete paginated Old School RuneScape Wiki shop listing for an item; get_item_sources is the bounded overview.
Inputs: item (required), limit, offset. Returns: item, limit, shops, total, +11 more
Get normalized facts from an Old School RuneScape Wiki item page; use get_item_sources for acquisition data.
Inputs: item (required). Returns: name, title, value, weight, +14 more
Get a bounded Old School RuneScape Wiki acquisition overview; follow truncation warnings with find_shop or find_drop_sources.
Inputs: item (required), perCategoryLimit. Returns: item, drops, shops, recipes, +5 more
Get separate Old School RuneScape Wiki monster variants and facts without combining variants or calculating DPS.
Inputs: monster (required), variant. Returns: monster, variants, warnings, mapPoints, +6 more
Get Old School RuneScape Wiki quest requirements without evaluating any player's progress.
Inputs: quest (required). Returns: items, quest, skills, warnings, +5 more
Get a bounded Old School RuneScape Wiki page. If truncated, use get_wiki_sections and get_wiki_section.
Inputs: title (required). Returns: title, content, sections, warnings, +3 more
Get one bounded Old School RuneScape Wiki section by the numeric index returned by get_wiki_sections.
Inputs: title (required), section (required). Returns: title, content, section, warnings, +3 more
List bounded section metadata for an Old School RuneScape Wiki page before calling get_wiki_section.
Inputs: title (required). Returns: title, total, returned, sections, +3 more
Search the Old School RuneScape Wiki for canonical article titles and snippets.
Inputs: query (required), limit, offset. Returns: total, offset, results, nextOffset, +1 more
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 10 tools
Each tool targets a distinct aspect of the Old School RuneScape Wiki: drops vs shops, item info vs sources, monster info, quest requirements, wiki page browsing (page, sections, section), and search. No overlap is apparent.
All tools follow a consistent verb_noun pattern using snake_case (e.g., find_drop_sources, get_item_info, search_wiki). There are no deviations or mixed conventions.
10 tools cover the main functionalities needed to interact with the OSRS Wiki: item lookup, monster info, quest requirements, wiki page retrieval and search. This is well-scoped and neither too many nor too few.
The tool set covers core wiki operations: searching, reading pages and sections, and extracting data on items, monsters, and quests. Minor gaps exist (e.g., no direct quest detail tool beyond requirements, no equipment info), but the surface is largely complete for a read-only wiki MCP.
Read-only MCP for the Eco game wiki: search, Markdown pages, and wiki_* lookups. No keys, no writes.
Read-only RS3 GE guide prices/history, optional reported buy/sell trades, buy-limit/alchemy facts, flip math and public hiscores. Search Wiki sources, list category members, and read structured monster/quest/drop facts. Guide prices lag activity; trades are not executable offers. Wiki article text requires reviewed revisions; other pages return links. Streamable HTTP; no account or API key. Independent Birdwell Labs project, unaffiliated with Jagex or the RuneScape Wiki.
Wikidata MCP — wraps Wikidata API (wikidata.org/w/api.php)
Wikipedia MCP — wraps Wikipedia REST API (free, no auth)
