Wirdi MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| WIRDI_AI_ROOT | No | Optional. Path of a Wirdi AI checkout: the server then reads and verifies the original files instead of the bundled snapshot. | |
| WIRDI_QURAN_SOURCE | No | Optional. Set to `none` to disable the Quran source: tools that read it return DATA_UNAVAILABLE and health_check shows quran_data: not_configured. | |
| WIRDI_MCP_LOG_LEVEL | No | Log level, written to stderr. | |
| WIRDI_KNOWLEDGE_ROOT | No | Optional. Path to another knowledge/ directory (default: this repository's knowledge/ directory). | |
| WIRDI_TEST_SOURCE_ROOT | No | Tests only: enables the real-corpus tests by pointing to a Wirdi AI checkout. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
Read-only Qaloun access and source mapping use checksum-pinned files (a bundled snapshot by default). Review-required mappings are unsafe for retrieval. Tajwid rules, candidate annotations and source-page lookup are local-only and unreviewed: no annotation is scholarly verified Corpus tariq is unspecified, not independently attested.
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
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| health_checkA | Report process readiness and configuration, without accessing source or private data. |
| get_qaloun_passageB | Read unchanged Qaloun ayat and verified page spans from checksum-pinned source files. |
| resolve_source_mappingC | Reuse the pinned mapper; review-required ranges are unsafe for source retrieval. |
| get_tajwid_ruleB | Read a local Tajwid rule record with its page-level citations (unreviewed). |
| list_tajwid_rulesB | List the available local rules and whether a detector exists for each. |
| analyze_qaloun_tajwidB | Candidate Tajwid occurrences for 1 to 100 Qaloun ayat, paged, each with its citations. |
| list_reading_combinationsD | Which readings of the madds may be taken together in each ayah (the book's tables). |
| summarize_tajwidC | Per-rule candidate counts for a surah or an ayah range, with per-ayah counts (paged). |
| analyze_app_pageC | Candidates for the ayat that begin on a page of the verified 559-page Qaloun layout. |
| get_source_pageA | Read 1 to 5 pages of a locally placed source book (printed page numbers). |
| search_source_pagesB | Literal Arabic search over locally placed source pages; returns page hits only. |
| get_letter_propertiesB | Place of articulation and characteristics of 1 to 29 Arabic letters (Book A pp.40-45). |
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 12 tools
Most tools target clearly distinct resources: passage retrieval, rule lookup, analysis, summarization, source pages, letter properties. The main potential confusion is between analyze_qaloun_tajwid and analyze_app_page (both return candidate occurrences, differing only in scope), and get_qaloun_passage vs get_source_page, but the descriptions clarify the boundaries.
The set follows a consistent verb_noun snake_case pattern (get_*, list_*, analyze_*, summarize_*, search_*, resolve_*). health_check is the sole outlier using a noun_verb form, though it is a widely recognized convention.
12 tools is a well-scoped surface for a specialized Tajwid/Qaloun analysis server, with retrieval, analysis, summarization, and diagnostic operations each earning a place.
The surface covers passage retrieval, rule listing/lookup, candidate analysis, summarization, source page access/search, letter properties, and reading combinations. Minor gaps exist, such as no explicit surah/ayah metadata listing or a way to search within Tajwid rule records, but core workflows are covered.