rei-meta-mcp
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., "@rei-meta-mcp Check coherence of SEED_KERNEL across all sources."
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.
rei-meta-mcp
A meta-layer standing on top of the connector group. It treats individual MCP connectors as objects and access paths to those objects as morphisms, and handles their structure.
The only practical purpose of Phase 1 is coherence check — mechanically cross-checking whether the contents match across multiple sources pointing to the same object.
Why it was made
On 2026-08-19, during the implementation of rei-memory-mcp, it was discovered that SEED_KERNEL had drifted by 1,677 vs 1,675 for 11 days without being noticed. It was noticed only because a human happened to compare both numbers by chance; there was no mechanical means of detection.
The first job of this connector is to let the machine find this first. See docs/incident-2026-08-19.md for details.
Related MCP server: MCP Discovery Server
Provided tools (3)
meta_list_sources(object_name?: str)
Enumerates registered sources and returns each one's reachability status, count, and git HEAD hash (a clue to freshness).
meta_check_coherence(object_name: str, detail: bool = False)
Cross-checks the fingerprints of sources pointing to the same object and returns one of the following verdicts:
verdict | meaning |
| Fingerprints of all reachable sources match |
| Mismatch among reachable sources |
| Zero reachable sources |
| Only one reachable source (not comparable) |
§4 Core rule: "unreachable" is not "matched". unreachable and single_source are always surfaced as warnings.
With detail=True, returns the actual list of mismatched IDs (up to 100 items).
meta_compose(from_source: str, to_source: str)
Phase 1 only does string matching of output_schema / input_schema in the registry. Actual schema inference is Phase 3 and later.
What Phase 1 does not do
Automatic repair (intentionally excluded. Judgment of which is correct is left to humans)
Persistence of inspection history (Phase 2)
Schema inference (Phase 3)
Categorical constructions such as functors, adjunctions, monads, etc. (when the need arises)
Currently registered sources
See config/sources.example.yaml:
source | kind | status |
|
| full fingerprint ( |
|
| partial fingerprint (launches |
|
| Deployment via claude.ai remote-devices cannot be probed directly from Python — always explicitly reports |
Install & usage
uv pip install -e ".[dev]"
cp config/sources.example.yaml config/sources.yaml # パスを埋める
uv run pytest # 全 PASS を確認
uv run rei-meta-mcp # stdio で MCP server 起動The registry path can be overridden with the environment variable REI_META_MCP_REGISTRY.
Honest scope
Detection only, no repair — does not take judgment outside of humans
Matching among partial fingerprint sources (mcp_stdio) is confirmation that "there is no mismatch", not proof of "exact match" (there are parts not visible at the
content_hashlevel)Only 5 category theory terms carry weight (object, morphism, equalizer, composition, identity morphism). Others are avoided
Currently a Phase 1 spike with 3 sources. A structure whose value increases as the number of objects grows
Phase 2 (history, scheduled execution, notifications) and Phase 3 (schema inference, generalization of morphisms) will be decided after Phase 1 works in real operation
License
AGPL-3.0-or-later.
Related
docs/incident-2026-08-19.md— record of the incidenttests/test_incident_2026_08_19.py— test that reproduces the incident
Available Tools
3 toolsmeta_check_coherenceB
Check whether all sources for object_name agree.
Verdicts: coherent | divergent | unreachable | single_source. §4: unreachable/single_source are warnings, not silent success.
Inputs: object_name (required), detail. Returns: object
meta_composeA
Check whether the output of from_source can feed to_source.
Phase 1: declarative schema string match only.
Inputs: to_source (required), from_source (required). Returns: object
meta_list_sourcesA
List registered sources with current reachability and freshness.
Inputs: object_name. Returns: object
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 3 tools
Each tool has a distinct responsibility: listing registered sources, checking source agreement for an object, and validating source-to-source composability. There is minimal overlap, and the differing parameters make selection unambiguous.
All tools use the `meta_` prefix with an imperative snake_case verb (`list`, `check`, `compose`), giving a predictable convention. The slight variation in whether an object follows the verb does not hurt recognizability.
Three tools is a compact but appropriate scope for a focused metadata validation server. Each tool provides a distinct high-level capability with no redundancy.
The set covers the core discovery and validation workflows: enumerate sources, check coherence, and test composition. It is missing broader source/object management and only performs schema-string composition matching, so there are minor gaps but no dead ends for the main use case.
Maintenance
Related MCP Connectors
One MCP URL for all your connectors — scoped writes, enforced constraints, and a full audit trail.
An MCP server that provides read access to your cloud storage providers, bank accounts and more.
MCP server for secureFlows: token-free URL builders and integration-linting tools for AI agents.
Single-connection meta-router indexing 500+ MCP servers & A2A tools with 85% token compression an.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceMCP server for automated architectural mapping, security vulnerability detection, ML asset tracking, and code metrics in local repositories.-
- FlicenseNot gradedqualityDmaintenanceA meta-MCP server that discovers and explores other MCP servers using progressive disclosure patterns to reduce token usage by up to 99%.-
- AlicenseNot gradedqualityDmaintenanceA comprehensive MCP server that provides a graph database for tracking software codebase components, their relationships, and associated tasks/goals.1MIT
- AlicenseNot gradedqualityDmaintenanceMCP server that enables coding agents to check documentation claims against source code, detecting drift and suggesting fixes.15 npm2MIT