@miraigent/free-ai-ops-mcp 0.1.21 → 0.1.23

package/CHANGELOG.md

@@ -1,5 +1,21 @@
 # Changelog
 
+## [0.1.23] - 2026-06-24
+
+### Changed
+
+- Added README guidance for reading returned MCP status fields so npm and
+  GitHub users can choose the next safe action before changing client config or
+  connecting private workflows.
+
+## [0.1.22] - 2026-06-22
+
+### Changed
+
+- Added response-reading guidance to the MCP JSON-RPC session example so npm and
+  GitHub users can tell whether the package is running before debugging desktop
+  MCP client configuration.
+
 ## [0.1.21] - 2026-06-21
 
 ### Changed

package/README.md

@@ -120,6 +120,38 @@ you ran, the public-safe input shape, and the returned status field. Do not
 paste private customer records, secrets, internal policy text, or paid product
 files.
 
+## Read The Returned Status
+
+After the smoke test runs, look for the status field from the tool you called
+before changing client configuration or private workflows.
+
+For `human_review_gate`:
+
+- `auto_ok`: the synthetic draft looks low risk, but the tool still does not
+  send anything.
+- `review_required`: keep a human reviewer in the workflow before sending.
+- `stop`: pause the workflow and resolve the sensitive, payment, complaint,
+  privacy, public-facing, or policy issue first.
+
+For `prompt_risk_review`:
+
+- `review_checklist_required`: document the data-handling checklist before
+  running the prompt.
+- `human_review_required`: add a human review step before AI output can affect a
+  customer-facing workflow.
+- `stop_before_ai_use`: do not run the prompt until the data-handling rule is
+  reviewed.
+
+For `faq_candidate_review`, use `recommendedStatus` to decide whether the
+pattern belongs in a public FAQ, internal FAQ, or human-review rule. For
+`ai_safe_crm_note`, check that facts, AI suggestions, human decisions, and next
+actions stay in separate fields.
+
+If the status is surprising, open a Tried It feedback issue with synthetic input
+only. The most useful report says which status you expected, which status you
+received, and what extra rule or explanation would have made the result easier
+to use.
+
 ## Copy A Useful Feedback Result
 
 When you open a Tried It feedback issue, the most useful report is the smallest

package/examples/mcp-json-rpc-session/README.md

@@ -23,6 +23,23 @@ include a human_review_gate result with:
 - reviewOwner: support lead
 - boundary: this tool is a review helper and does not send messages
 
+## Read The Responses
+
+Use the response shape as a quick acceptance check before changing a desktop MCP
+client config:
+
+- Response 1 should include `serverInfo.name` set to
+  `miraigent-free-ai-ops-mcp`.
+- Response 2 should include four tool names:
+  `human_review_gate`, `faq_candidate_review`, `ai_safe_crm_note`, and
+  `prompt_risk_review`.
+- Response 3 should return a `human_review_gate` content block whose JSON text
+  includes `gateStatus`, `reviewOwner`, `nextAction`, and `boundary`.
+
+If those fields appear in the terminal but not in Claude Desktop, Cursor, or
+another client, the package is likely running and the next thing to check is the
+client config, restart behavior, or command path.
+
 ## Public-Safe Feedback
 
 If the output is unclear, open a Tried It feedback issue and include only:

package/mcp/free-ai-ops-server.mjs

@@ -1,6 +1,6 @@
 #!/usr/bin/env node
 
-const serverInfo = { name: 'miraigent-free-ai-ops-mcp', version: '0.1.21' };
+const serverInfo = { name: 'miraigent-free-ai-ops-mcp', version: '0.1.23' };
 
 const tools = [
   {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@miraigent/free-ai-ops-mcp",
-  "version": "0.1.21",
+  "version": "0.1.23",
   "description": "Free MCP server for developers adding human review gates, prompt risk checks, FAQ review, and CRM note safety to AI tools.",
   "homepage": "https://github.com/Miraigent/miraigent-free-ai-ops-mcp",
   "type": "module",