shared-wiki-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| WIKI_VAULT | No | Path to the wiki vault. Defaults to 'vault' in the current directory. Can also be set via the --vault command-line option. | vault |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| wiki_searchB | Search note titles and content; return ranked excerpts and status. |
| wiki_readB | Read body lines, metadata, hash, links, parents and backlinks of a note. |
| wiki_contextB | Gather bounded source excerpts for a question. The host composes the answer. |
| wiki_review_queueB | Rank open questions by a transparent heuristic of impact, uncertainty and time. |
| wiki_changesB | List notes updated since a Unix millisecond timestamp. Not a semantic diff. |
| wiki_captureB | Save user feedback as an unreviewed inbox note; source_ref is provenance, not authenticated identity. |
| wiki_proposeB | Create a separate proposed replacement after reading the current target. Never overwrites it. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 7 tools
Each tool targets a distinct operation: search for discovery, read for a single note's body/metadata, context for bounded multi-source excerpts, changes for a time-based delta, capture for inbox writes, propose for replacement drafts, review_queue for ranking open questions. The only soft overlap is wiki_search vs wiki_context (both return excerpts), but the descriptions clarify that one is ranked retrieval and the other is question-scoped gathering.
All seven tools share the wiki_ prefix and snake_case, giving a predictable namespace. However, the second element mixes verbs (search, read, capture, propose) with nouns (review_queue, changes, context), so the verb_noun pattern is not uniformly applied.
Seven tools is a tight, well-scoped surface for a shared wiki: retrieval, inspection, review, change tracking, and two write paths. Every tool maps to a real workflow step with no filler.
Read/discovery/write-capture are covered, but two lifecycle paths dead-end: wiki_propose creates a proposed replacement with no tool to list, accept, or reject proposals, and wiki_review_queue surfaces open questions with no tool to answer or resolve them. Agents can work around this by reading and capturing notes, but the review/proposal loop is incomplete.