# Vibe Monitor [Verified] [Health: Active]

**Category:** 💻 Developer Tools  
**Repository:** https://github.com/tubban1/vibe-monitor  
**GitHub Stars:** 4  
**Views:** 4  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/vibe-monitor

## Description
Vibe Monitor gives AI coding agents session-scoped architecture checkpoints, historical signals, and practical next steps for growing projects. Read-only source analysis runs locally; source files, diffs, prompts, project paths, and secrets are not uploaded to Vibe Monitor servers. Works with Cursor, Claude Code, Codex, Trae, and other supported agent integrations; WorkBuddy support is experimental. Requires Node.js 22+. Free analysis uses a cloud policy service and default-on bounded pseudonymous telemetry. Install and runtime commands follow @latest. The public client is Apache-2.0 licensed.

## Tools
Capabilities this server exposes over MCP:

- **vibe_version** — Get the Vibe Monitor version currently running in this MCP process, the installed current/previous runtime versions, staged update, and update state. A process may still run an older bundle after the installed pointer changes. Use this tool when the user asks which Vibe Monitor version is running or installed. schemaVersion and evidenceVersion are data format versions, NOT the software version.
- **vibe_account_connect** — Explain how to link this Vibe Monitor installation to the user's account. Call ONLY when the human asks to link this device. Makes no network request.
- **vibe_session_start** — Start or resume Vibe Monitor for the current coding session. Can be interrupted or explicitly skipped without preventing coding. For an explicit user Skip request, set mode=skip. Otherwise establishes a bounded project assessment.
- **vibe_skip_current** — Skip an in-progress Vibe Monitor Session Start, at the user's explicit request. Cancels active scanning in this MCP server without stopping coding or disabling future monitoring. Agent hosts may expose this tool as an action; Vibe Monitor cannot insert custom buttons in host UI.
- **vibe_checkpoint** — Checkpoint structural evolution after a meaningful implementation batch. Persists a snapshot and reports stage-aware findings.
- **vibe_project_status** — Read current architecture-evolution status without creating a history snapshot.
- **vibe_findings** — Return current findings ordered by evolution stage and pressure score. Findings include Do now, Prepare, Do not do yet, and Next trigger.
- **vibe_finding_intelligence** — Read risk trend, evidence-review status and explicitly confirmed Product Intent guidance for one Finding. Includes an unverified baseline for future feedback, never causal attribution.
- **vibe_explain_finding** — Explain one finding with evidence, standards mapping, heuristic disclosure, staged action and over-engineering guard.
- **vibe_product_intent** — Read Vibe Monitor's current product-intent context. Inferred values are advisory; explicit values come only from the user's own stated intent.
- **vibe_set_product_intent** — Store explicit product-intent facts locally when the user has clearly stated them. Do not turn guesses into explicit intent and do not ask these questions unless an architecture decision materially depends on them.
- **vibe_get_architecture_review_request** — Create a bounded ARP v1 semantic architecture review request for a finding that explicitly recommends semantic review. The request contains normalized architecture evidence and no raw source code or project path.
- **vibe_submit_architecture_review** — Submit one ARP v1 semantic architecture review. Different agents may legitimately return different verdicts. Full model-generated review text stays local; Cloud receives only bounded structured review metadata for comparison and outcome learning.
- **vibe_outcome_record_action** — ONLY after a user or agent explicitly states they applied, deferred, or dismissed a specific Vibe recommendation. Record a REPORTED action, not inferred code changes; never claim a fix occurred. Idempotent via stable submissionId.
- **vibe_outcome_verify** — Record ONLY explicit user-reported verification for a previously reported action. User affirmation plus regression-test/independent-review evidence type is still not proof of causation. Never fabricate execution or user confirmation.
- **vibe_outcome_status** — Read local recommendation→reported action→comparable follow-up→user verification and recurrence history. No project scan or Cloud telemetry. Risk improvement is observational, attribution remains unproven.
- **vibe_feedback** — Record explicit user feedback about a Vibe Monitor finding. Use only when the user clearly says a warning was useful, wrong, too early, too late, or confusing. Never ask for feedback after every warning.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "vibe-monitor": {
    "command": "npx",
    "args": ["-y","--registry=https://registry.npmjs.org","@vibe-monitor/setup@latest"]
  }
}
```

## Documentation

## What Vibe Monitor does

Vibe Monitor acts as a local architecture advisor for AI-assisted coding workflows. It examines accessible project code without modifying it, then helps an agent identify accumulating architectural pressure, coupling, module-responsibility issues, and testing signals. Recommendations are organized around timing: what to do now, what to prepare for, what not to address yet, and which future change should trigger another review.

The system is session-oriented rather than an always-running daemon. A session can be started or resumed before work, skipped at the user's request, and checkpointed after a meaningful implementation batch. Historical snapshots make it possible to compare how architectural signals change across coding sessions.

## How it works

Structural scans run locally and do not call an LLM. Optional agent-assisted interpretation can use model tokens, while the local client retains the full review text for semantic architecture reviews. Vibe Monitor does not upload source files, diffs, prompts, project paths, or secrets to its servers. Derived metrics and product-operation events may be sent through pseudonymous telemetry, and free architecture analysis uses a cloud policy service.

Findings expose an evolution stage and pressure score. Agents can request an explanation for an individual finding, including its evidence, standards mapping, heuristic disclosure, staged action, and an over-engineering guard. Product-intent context can distinguish advisory inferences from facts explicitly stated by the user. Explicit intent is stored locally when the user has clearly provided it.

For findings that recommend semantic review, the client can create a bounded ARP v1 request containing normalized architecture evidence rather than raw source code or a project path. A submitted review sends bounded structured metadata for comparison and outcome learning; full model-generated review text remains local. Different agents may return different verdicts.

## Setup and configuration

Install the unified open-source client with Node.js 22 or newer:

```bash
npx -y --registry=https://registry.npmjs.org @vibe-monitor/setup@latest install
```

The installer registers supported agent MCP or rules configuration and may create local history or backups. No activation code is required, and no separate monitoring daemon needs to be installed. The same client is used across Free, Pro, and Team plans.

New installations enable bounded pseudonymous product telemetry by default with notice. Use `install --no-telemetry` or run `vibe-monitor telemetry off` to disable collection. Disabling telemetry pauses Free architecture analysis, and an existing opt-out remains in effect after reinstall. Account linking is available through the MCP interface when the user explicitly requests it; the account-connect operation itself makes no network request.

## Tools and capabilities

The MCP tool set covers the complete review lifecycle:

- `vibe_session_start` starts or resumes a bounded assessment, while `vibe_skip_current` cancels an in-progress start without stopping coding or future monitoring.
- `vibe_checkpoint` stores a snapshot after an implementation batch. `vibe_project_status` reads current architecture-evolution status without creating a snapshot.
- `vibe_findings` returns stage-ordered findings and pressure scores. `vibe_explain_finding` provides evidence and a staged explanation, while `vibe_finding_intelligence` reports risk trends and review status.
- `vibe_product_intent` reads intent context, and `vibe_set_product_intent` stores facts the user has explicitly stated.
- `vibe_get_architecture_review_request` prepares bounded semantic-review input, and `vibe_submit_architecture_review` submits one review.
- `vibe_outcome_record_action`, `vibe_outcome_verify`, and `vibe_outcome_status` track explicitly reported actions, verification, comparable follow-up, and recurrence. They do not infer that code changed or that a recommendation caused an improvement.
- `vibe_feedback` records clear user feedback when a finding was useful, wrong, premature, late, or confusing.
- `vibe_version` reports the running process version, installed versions, staged update, and update state. `vibe_account_connect` explains how to link an installation when requested.

The public client is Apache-2.0 licensed and auditable. Cloud policy and scoring configuration is a separate proprietary service; its delivery does not authorize raw-source uploads or arbitrary remote-code execution.

_Full upstream README: https://allmcps.com/mcp/vibe-monitor/readme_