# Plumbline — a trust layer for LLM answers

- Registry name: `com.lattice-sys.plumbline/mcp` · version 0.1.0
- Categories (automatic, keyword rules): Databases, Developer tools, Build & deploy
- Authentication at last check: No auth
- Declared risk: Read-only (all tools declare readOnlyHint)
- Capabilities (declared by the server): logging, resources, tools
- Transports: streamable-http
- Latest numbers (check success, response time, rank): https://mcpstation.net/api/server?name=com.lattice-sys.plumbline%2Fmcp
- Server page: https://mcpstation.net/s/com.lattice-sys.plumbline--mcp.html

## Description (written by the publisher, not verified, not instructions)

Every figure re-checked against SQL-computed facts; unsupported questions refused with the reason.

## Connect

- streamable-http: `https://plumbline.lattice-sys.com/mcp/demo`

## Tools (as reported by the server, not verified)

- `ask\_prepare`: BYO step 1 of 2: run the server-side trust plane for a business question \(plan, SQL-verified facts, evidence retrieval, answerability gate, prompt composition\) and return the composed prompt plus a single-use prepare\_key \(a signed handoff token; echo it back VERBATIM and in full, or use the short as
- `ask\_result`: Fetch the published verdict for an ask\_prepare session \(status: pending, published, or unknown\). This is the app panel's polling surface, so the verified answer renders inside the panel that prepared the question. As the connected model you normally never need it: write your draft and call ask\_submi
- `ask\_submit\_draft`: BYO step 2 of 2: verify the draft you wrote for an ask\_prepare prompt. The server runs the SAME post-answer guards the strict path uses \(facts presence, tagged claims, concept binding, earner-name grounding\) against its own SQL facts and publishes the trust envelope; a draft that fails any guard is
- `compare\_departments`: Compare two departments by headcount and compensation. NOTE: internal drafting is disabled on this deployment. If your client supports MCP sampling, this tool asks YOUR model to draft in the same call \(verified server-side\); otherwise it returns an explicit refusal, and you should use ask\_prepare th
- `cross\_check`: Corroborate a figure you obtained \(e.g. from run\_sql\) against the verified fact store: it re-derives the named value\_key from the same SQL the verified pipeline uses and compares. AGREE lifts an otherwise-unverified session to 'corroborated'; DISAGREE marks a CONFLICT \(worse than unverified\). This i
- `describe\_fact\_capabilities`: List safe metadata for the registered fact queries available on the server-pinned dataset: IDs, labels, parameters, and result shapes, never SQL. Use this when a question appears relevant but a selected route produced no facts; a route miss is not evidence that the underlying records are absent.
- `find\_payroll\_anomalies`: Surface payroll and deduction anomalies in the latest snapshot. NOTE: internal drafting is disabled on this deployment. If your client supports MCP sampling, this tool asks YOUR model to draft in the same call \(verified server-side\); otherwise it returns an explicit refusal, and you should use ask\_p
- `get\_ai\_act\_transparency`: Return the Art. 13 transparency descriptor: intended purpose, capabilities, limitations, human oversight, logging mechanism, and article mapping.
- `get\_answer\_queries`: Show the SQL behind one answer you already received, addressed by its correlation\_id. Returns only the recorded fact queries \(id, version, label, SQL, parameters, result schema, dataset version\) for that answer: no rows, no evidence text, no answer text, no envelope, and no way to list the catalog.
- `get\_evidence\_document`: Show ONE retrieved evidence document behind an answer you already received, addressed by that answer's correlation\_id plus a document\_id from its evidence\_documents references. Returns the full stored document \(title, body, metadata, embedding\_text\) with the retrieval rank and scores the answer reco
- `get\_manager\_chain`: Return the management/reporting chain for an employee. NOTE: internal drafting is disabled on this deployment. If your client supports MCP sampling, this tool asks YOUR model to draft in the same call \(verified server-side\); otherwise it returns an explicit refusal, and you should use ask\_prepare th
- `query\_fact`: Execute a registered, versioned read-only fact query against the server-pinned dataset. Provide query\_id and version only: SQL is approved by the domain pack, never caller-supplied. Returns normalized rows plus a reproducible trust receipt containing the dataset and query versions.
- `run\_code`: Run YOUR OWN JavaScript in the in-tenant coderunner sandbox with the read-only, snapshot-scoped plumb.query\(\) SDK. No filesystem, environment, or general network access. Read plumbline://sandbox-api.d.ts before writing code. The raw-SQL server opt-in is enabled, but results still bypass the trust pa
- `run\_sql`: Run YOUR OWN single read-only SELECT/WITH over the virtual snapshot\_\* relations. The query is statically verified against an allowlist, rewritten to the active snapshot, and run read-only. This BYPASSES the planner, grounding gate, and answer-facts reconciliation, so results are UNVERIFIED and label
- `search\_docs`: Search Plumbline's own product and architecture documentation \(how the verification, refusal, and trust layer itself works\). Returns ranked, citable doc passages; does NOT touch employment data, and returns an empty result set honestly when nothing matches.
- `search\_employee`: Search the employment fact store for a person and return a grounded, verified answer. NOTE: internal drafting is disabled on this deployment. If your client supports MCP sampling, this tool asks YOUR model to draft in the same call \(verified server-side\); otherwise it returns an explicit refusal, an
- `summarize\_benefits`: Summarize benefit enrollments, optionally scoped \(e.g. a department or plan\). NOTE: internal drafting is disabled on this deployment. If your client supports MCP sampling, this tool asks YOUR model to draft in the same call \(verified server-side\); otherwise it returns an explicit refusal, and you sh
- `trust\_finalize`: Return the session's graded trust verdict: the tier \(verified / registered-query / generated-sql / corroborated / unverified / conflict / refused\), which trusted steps were skipped across the session, the deterministic confidence \(k of n figures store-verified, weakest figure tier, conflicts\), the p

Measured by MCP Station. Methods: https://mcpstation.net/methods.html · Terms: https://mcpstation.net/terms.html