# Zorqelio — Data Reliability for AI Agents

> MCP policy creation, quality checks and guarded data handoffs before automated writes.

- name: io.github.ssitdeveloper/zorqelio
- version: 1.3.0
- connection class: R0 (autonomous)
- trust: 97/100
- quality: 84/100 (new)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/io.github.ssitdeveloper%2Fzorqelio
- tools: https://api.protogrid.dev/v1/servers/io.github.ssitdeveloper%2Fzorqelio/tools

**An agent can connect right now, no human step.** Remote endpoint, no authentication, reachable on the last probe.

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "zorqelio": {
      "type": "http",
      "url": "https://zorqelio.xyz/mcp"
    }
  }
}
```


## Trust

- hygiene: 100
- liveness: 100
- freshness: 100
- provenance: 90
- drivers: +repository +namespace-matches-repo +website +reachable +uptime-100% +updated-0d-ago
- Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed.

## Quality 84/100 (new)

### protocol: 88 (weight 25)

- pass `protocol.modern`: supports 2026-07-28 (server/discover)
- pass `protocol.stateless`: answers without a session
- pass `protocol.transport`: streamable HTTP
- warn `protocol.list_ttl`: tools/list declares ttlMs 0, so clients re-fetch it on every use

### authorization: n/a (weight 15)

- n/a `auth.prm`, `auth.as_metadata`, `auth.cimd`: no remote uses OAuth
- n/a `auth.secret_in_url`: no templated URL

### tool hygiene: 81 (weight 30)

- pass `tools.descriptions`: every tool has a description
- pass `tools.description_length`: descriptions are concise
- pass `tools.schemas`: every tool has an object input schema
- warn `tools.annotations`: only 8 of 29 tools declare readOnlyHint or destructiveHint
- fail `tools.directory_hints`: no tool declares all four of readOnlyHint, destructiveHint, idempotentHint and openWorldHint; missing: data_quality_check (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); quality_policy_publish (openWorldHint); workspace_create (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); …
- pass `tools.token_cost`: about 4,424 tokens to load every tool
- pass `tools.api_dump`: tools are not a one-to-one API dump
- pass `tools.hidden_text`: no hidden text in tool names, descriptions, schemas or annotations

### stability: n/a (weight 15)

- n/a `stability.changes`, `stability.rug_pull`, `stability.new_links`: tool history recorded for 1 of the 7 days needed

### dependencies: n/a (weight 15)

- n/a `deps.known_vulns`, `deps.mcp_sdk_version`, `deps.resolvable`: no npm or PyPI package

- tools: 29, about 4,424 tokens to load them all
- tool set hash: 4b1db5ab0e176e9c2452eb1aa21f822a
- tools last changed: 2026-11-13T23:47:21.462Z
- badge: [![protogrid quality](https://protogrid.dev/badge/io.github.ssitdeveloper/zorqelio.svg)](https://protogrid.dev/servers/io.github.ssitdeveloper/zorqelio)
- Checks run on what our credential-free, read-only probes observe; tools are never called and no code audit is performed.

### Tool changes

- 2026-11-13 `handoff_evaluate` changed (inputSchema)
- 2026-11-13 `quality_policy_evaluate` changed (inputSchema)
- 2026-11-13 `quality_policy_publish` added
  - Publish a new immutable policy version. Pin expectedCurrentVersion to prevent overwriting a concurrent update. Publishing invalidates unclaimed handoffs pinned to the old version. Reuse operationId only for exact retries.
- 2026-11-13 `quality_policy_create` added
  - Create an owned, versioned quality policy for an existing workspace contract. Requires active subscribed or evaluation access. Pass rules as JSON text and a stable operationId; exact retries return the same policy without creating duplicates.
- 2026-11-13 `handoff_history` added
  - List recent private guarded data handoffs with sanitized failure reasons. No input payloads are stored.
- 2026-11-13 `handoff_get` added
  - Inspect a private guarded handoff decision and its customer-reported write outcome. Reading never claims or renews a permission.
- 2026-11-13 `handoff_evaluate` added
  - Check the exact payload against pinned schema, quality and optional source-readiness rules. Returns a 60-second decision. Does not execute a write; use the guarded SDK integration to claim permission once. Reuse operationId only for identical input.
- 2026-11-13 `quality_policy_history` added
  - Read recent private evaluations of one saved quality policy.
- 2026-11-13 `quality_policy_evaluate` added
  - Evaluate one explicit saved policy version, persist sanitized results and return ALLOW, WARN or BLOCK. Reuse operationId for an exact retry.
- 2026-11-13 `quality_policy_get` added
  - Retrieve a saved private data-quality policy version.

Full history: https://api.protogrid.dev/v1/servers/io.github.ssitdeveloper%2Fzorqelio/changes

## Remotes

- https://zorqelio.xyz/mcp (streamable-http, auth none, reachable true, uptime30d 1)

## Packages

_none_

## Tools (29)

- `contract_compare`: Compare two immutable versions of a private data contract and report added, removed, and changed schema paths. Useful for reviewing upstream API contract changes before publishing or upgrading an agent.
- `contract_create`: Create a private, reusable, versioned data contract for an AI-agent workflow. Returns a one-time access token; store it securely because it cannot be retrieved later.
- `contract_get`: Retrieve metadata and a published schema version for a private data contract. Requires its access token.
- `contract_history`: Retrieve recent validation decisions and failure evidence for a private contract. Stores only validation metadata, not customer payloads; the latest 100 events per contract are retained.
- `contract_publish`: Publish a new immutable version of an existing private data contract. Older versions remain available for reproducible validation.
- `contract_validate`: Validate incoming JSON against a stored, versioned data contract. Detects missing, changed, or invalid upstream data before an AI agent uses it. Returns machine-readable pass/fail and evidence.
- `data_normalize`: Normalize inconsistent JSON records into a predictable structure using explicit source-to-target field mappings and defaults. Deterministic; does not call an AI model.
- `data_quality_check`: Run deterministic data-quality rules against one JSON object or up to 10,000 records. Checks required/non-empty fields, types, numeric ranges, allowed values, email/URL formats, and uniqueness. Returns a weighted quality score and actionable row-level issues.
- `data_quality_profile`: Profile a JSON dataset and report field completeness, nulls, empty values, distinct counts, type distributions, and duplicate rows. Deterministic and local to the request.
- `handoff_evaluate`: Check the exact payload against pinned schema, quality and optional source-readiness rules. Returns a 60-second decision. Does not execute a write; use the guarded SDK integration to claim permission once. Reuse operationId only for identical input.
- `handoff_get`: Inspect a private guarded handoff decision and its customer-reported write outcome. Reading never claims or renews a permission.
- `handoff_history`: List recent private guarded data handoffs with sanitized failure reasons. No input payloads are stored.
- `json_diff`: Compare two JSON documents and return deterministic JSON Pointer paths that were added, removed, or changed.
- `json_format`: Pretty-print and normalize JSON. Accepts JSON text and returns formatted JSON plus basic statistics.
- `json_schema_validate`: Validate a JSON document against a supplied JSON Schema and return actionable validation errors.
- `json_to_csv`: Convert a JSON array of flat objects into CSV with escaped cells and a stable union of columns.
- `monitor_create`: Create a private push-monitor endpoint connected to an existing contract. Send each upstream API response to the returned HTTPS endpoint; Zorqelio validates it, records ALLOW/BLOCK evidence, and never stores the submitted payload. The monitor token is shown once.
- `monitor_history`: Retrieve the recent event timeline for a private monitor, including creation, stale transitions, recovery, blocked checks, and revocation. Requires its monitor ID and one-time monitor token. Event records never contain customer payloads.
- `monitor_revoke`: Immediately revoke a private push monitor token. Future submissions to its HTTP endpoint are rejected. Requires its monitor ID and monitor token.
- `monitor_status`: Retrieve health and usage counters for a private push monitor. Requires its monitor ID and one-time monitor token.
- `quality_policy_create`: Create an owned, versioned quality policy for an existing workspace contract. Requires active subscribed or evaluation access. Pass rules as JSON text and a stable operationId; exact retries return the same policy without creating duplicates.
- `quality_policy_evaluate`: Evaluate one explicit saved policy version, persist sanitized results and return ALLOW, WARN or BLOCK. Reuse operationId for an exact retry.
- `quality_policy_get`: Retrieve a saved private data-quality policy version.
- `quality_policy_history`: Read recent private evaluations of one saved quality policy.
- `quality_policy_publish`: Publish a new immutable policy version. Pin expectedCurrentVersion to prevent overwriting a concurrent update. Publishing invalidates unclaimed handoffs pinned to the old version. Reuse operationId only for exact retries.
- `workspace_create`: Create a private customer workspace. Returns a workspace API token once; save it securely.
- `workspace_list`: List contracts and aggregate usage belonging only to a private workspace. Requires its workspace API token.
- `workspace_revoke`: Permanently revoke a workspace API token and disable workspace-level operations. This action cannot be undone; existing individual contract and monitor tokens are not revoked.
- `workspace_token_rotate`: Rotate a workspace API token. The current token is required; the replacement is shown once and the old token stops working immediately.