# ai.geodesiclabs/governance-platform

- Registry name: `ai.geodesiclabs/governance-platform` · version 1.2.0
- Categories (automatic, keyword rules): AI models & agents, AI, media & content, Payments & accounting, Finance & markets, Health & fitness
- Authentication at last check: No auth
- Declared risk: Has destructive tools (destructiveHint)
- Capabilities (declared by the server): experimental, prompts, resources, tools
- Transports: streamable-http
- Latest numbers (check success, response time, rank): https://mcpstation.net/api/server?name=ai.geodesiclabs%2Fgovernance-platform
- Server page: https://mcpstation.net/s/ai.geodesiclabs--governance-platform.html

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

Pre-execution governance for AI agents. Deterministic PASS/FAIL/REVIEW verdicts, replayable proof.

## Connect

- streamable-http: `https://app.geodesiclabs.ai/mcp`

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

- `validate`: Validate structured data against a Blueprint's rules BEFORE the result is used. Returns PASS, FAIL, or REVIEW with plain-language findings, repair suggestions, a determinism hash, and a re-verifiable certificate. Same input + same rules = same verdict, every time.
- `validate\_repair`: Validate structured data against a Blueprint and, when it fails, include repair suggestions \(corrected values with the rule each fix is based on\) in the same call. Same verdicts as validate: PASS, FAIL, or REVIEW, with reasons and proof.
- `create\_blueprint`: Create a Blueprint - the governance contract validation runs against. A Blueprint defines what correct means for your data: fields, the math that must hold between them, and acceptable ranges. Start from load\_rule\_pack or discover\_patterns if you have no rules yet; invoke the blueprint\_guide prompt
- `list\_blueprints`: List the Blueprints on this account with field/rule/constraint counts and mode. Use the returned workflow\_name as 'blueprint' in validate.
- `repair`: One-shot repair: return corrected values that would make failing data valid under the Blueprint. Use repair\_path to see the steps instead.
- `check\_blueprint\_health`: Static pre-deploy analysis of a Blueprint's rule set. Returns a health verdict - healthy, acceptable, fragile, rigid, split, brittle\_islands, or unsatisfiable - with advice, including joint conflicts pairwise checks miss.
- `compare\_semantic\_equivalence`: Compare two payloads under the dual-hash design: content\_hash is content\_hash normalizes field order and numeric formatting. Semantic comparison preserves field roles; renaming requires an explicit bijection. With supplied rules, scalar types and whitespace remain significant. Structural similarity
- `govern\_inference`: Quality-govern an in-progress AI generation step BEFORE its output is used \(complements validate, which checks finished documents\). Returns an action - STOP, CONTINUE, REPAIR\_REGION, REUSE\_MOTIF, REVIEW, ESCALATE - with a plain-language explanation. Structural scores do not establish task correctnes
- `get\_inference\_trace`: Retrieve the durable audit trail for a governed generation: every recorded decision and its reasons.
- `recent\_inference\_decisions`: Recent generation-governance decisions in this owner's Blueprint version.
- `verify\_certificate`: Independently re-verify a validation certificate. Integrity mode checks the hash chain; full mode \(certificate + original data\) recomputes every attested rule from scratch - trust nothing, recheck everything.
- `profile\_blueprint\_robustness`: Sweep the Blueprint's numeric constraint bounds and report verdict stability: the stable band, the scales where the verdict first flips, and advice. Use before deploying bound changes.
- `forecast`: Deterministic forward reasoning: from the current data state, generate and rank the valid next states reachable under the Blueprint's rules.
- `discover\_patterns`: Learn candidate validation rules and structural document types from a batch of your records, deterministically - no Blueprint required. Promote results with approve\_rule. Source data is not stored.
- `repair\_path`: Find the shortest sequence of field changes taking invalid data to a valid state, as an ordered path of intermediate states. Different from repair \(one-shot nearest fix\): use repair\_path to explain or audit the fix, or compare alternative repairs.
- `counterfactual`: Run the same data under two rule sets and compare which future states remain valid - what-if analysis for rule changes.
- `analyze\_anomaly`: Explain whether a record fits the usual pattern for records like it, and which fields stand out. No Blueprint required.
- `create\_chain`: Create a multi-agent sequential chain: stages validate in order against one Blueprint, repairs propagate forward, TTL bounds the run. Siblings: submit\_chain\_stage advances the chain; handoff\_audit verifies a transition between stages. Returns chain\_id.
- `submit\_chain\_stage`: Submit data for the chain's current stage; the platform validates it and advances the chain if it passes. Response includes next-stage info and accumulated repairs.
- `handoff\_audit`: Audit a handoff between two chain stages: a context capsule of verified facts from the prior stage, and \(if proposed\_data is given\) a compatibility verdict that catches fields mutated in transit. Siblings: create\_chain, submit\_chain\_stage.
- `approve\_rule`: Promote a rule discovered by discover\_patterns into Blueprint-ready form.
- `reject\_rule`: Reject a discovered candidate rule so it will not be promoted into a Blueprint. Pair with approve\_rule after discover\_patterns.
- `structural\_types`: Retrieve the document categories a discover\_patterns session identified \(counts, distinguishing fields, domain hints\). Read-only; returns status=no\_session if discovery has not run for this namespace.
- `decompose\_failure`: Measure the numeric difference between original and corrected values: how much lies along the supplied linear rule directions \(projection\) and how much lies outside them \(residual\), with per-field contributions. Diagnostic only: it does not establish a cause, and neither record is verified. Use with
- `geometric\_confidence`: Summarize an already-computed state\_vector into a confidence level \(high/medium/low\) with a recommendation. Post-hoc digest - use analyze\_anomaly or check\_drift for fresh analysis of raw data.
- `check\_realization`: Structural realization analysis of a payload against the Blueprint's reference configuration \(requires a 'realization' block; otherwise status=skipped\). Diagnostics-tier tool; prefer validate or analyze\_anomaly for standard checks.
- `check\_drift`: Check whether recent submissions still match the established pattern for this Blueprint. Returns a stability verdict and observation count.
- `authorize\_execution`: Go/no-go for a real-world action \(payment, filing, API write\): runs full validation, then the Blueprint's execution gate. authorized=true only on PASS; REVIEW means do not proceed automatically. Different from validate: validate asks is this data correct, authorize\_execution asks should this action
- `load\_rule\_pack`: Load a prebuilt Blueprint template \(invoices, timecards, legal, POs, claims\). Call without pack\_id to list packs; then create\_blueprint to save a customized copy.
- `get\_execution\_trace`: Run validation and return the per-node execution trace \(node names, deterministic flags, timing\) plus the verdict and determinism hash. Use validate for normal operation; this is for debugging and audit preparation.
- `verify\_replay`: Check replay commitment integrity and compare recorded execution components. Version 4 includes reference context and the final result/status. A match compares commitments; this tool does not rerun the workflow or reconstruct historical reference populations, and does not prove factual correctness.
- `account\_status`: This account's plan, key usage, Blueprint counts, and the deployed platform build fingerprint \(version, build, deployed\).
- `delete\_blueprint`: Permanently delete a Blueprint and revoke its API keys. Irreversible; requires confirm=true. Account-level keys are unaffected.
- `update\_blueprint`: Update an existing Blueprint in place. Only passed fields change; pass \[\] to clear a list. workflow\_name cannot be renamed and existing API keys keep working. Different from create\_blueprint: modifies an existing Blueprint, mints no new key.
- `list\_api\_keys`: List this account's API keys \(masked\) with their Blueprint bindings.
- `rotate\_api\_key`: Replace an API key with a fresh one. The old key stops working immediately; the new key inherits its bindings.
- `delete\_api\_key`: Permanently delete one of the caller's API keys. DESTRUCTIVE — agents using the deleted key will receive auth errors immediately. The Blueprint a key was tied to \(if any\) is NOT affected; only the credential is revoked. To delete a Blueprint and all its keys, use delete\_blueprint. The target key can

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