codinfy-agent-monitor
Provides tools for checking Git status, including branch information, changes, and commit readiness, allowing agents to monitor version control state.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@codinfy-agent-monitorcheck agent status and usage limits"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Codinfy Agent Monitor
Real-time mission control for AI engineering teams.
Observe agents, context, limits, workflows, Git, security and model economy from one local-first command center.
Demo · v0.2.0 · Quick start · Command center · MCP API · Security · Contributing · License
◈ Live dashboard demo
Mission-control showcase
Live local interface — /codinfy
Related MCP server: WhenLabs/When
✨ What's new in v0.2.0
Area | v0.2.0 update |
Node Server Monitor | Cross-platform live inventory for Node/Bun/Deno servers, ports, framework, detected project, CPU, memory, uptime, orphan candidates, conflicts, and protected processes |
Safe process control | Live identity revalidation, explicit confirmation before stop, protected-process refusal, and Force Kill locked behind a failed graceful stop plus a second confirmation |
MCP Update Center | SemVer comparison with the official GitHub release, release notes, breaking changes, install-method detection, preflight, backup, confirmed npm update, and rollback |
Developer cockpit | Ten new routes: Node servers, ports, process map, resources, updates, release notes, backups, doctor, recovery, and notifications |
Continuity & health | Checksum-protected configuration backup/restore, MCP Health Doctor, read-only session recovery, Resource Guard, and local notification settings |
Expanded MCP/CLI | 66 MCP tools plus nested |
Read the complete v0.2.0 changelog · Open the design specifications
✦ The mission
AI agents move fast. Their context, tool calls, file changes, limits and failures usually remain scattered across terminals and provider interfaces. Codinfy Agent Monitor turns that invisible activity into a single operational picture.
Mission system | What it reveals | |
🛰️ | Agent Radar | Active, idle, thinking, running, blocked, failed and completed agents |
◫ | Context & Limits | Context, current rate, daily and weekly usage with official/estimated provenance |
⟁ | Workflow Pulse | Tasks, owners, progress, blockers, files and a real-time event timeline |
◈ | AI Credit Saver | Model Need Score, safer model category and estimated token/cost savings |
⛨ | Safe Guard | Secret scanning, sensitive-file detection and review before commit |
⎇ | Dev Operations | Git state, tests, builds, environment and public-ready checks |
◉ | Codix Observer | Context-aware recommendations for risk, economy and next actions |
⇄ | Universal MCP | Claude Code, Codex, Cursor, Windsurf and other MCP-compatible hosts |
Exact provider limits are displayed asofficial only when a host or adapter supplies them directly. Otherwise the interface clearly enables estimate mode. Codinfy Agent Monitor never presents a local estimate as provider truth.
⚡ Launch in 60 seconds
git clone https://github.com/bakalagoin/codinfy-agent-monitor.git
cd codinfy-agent-monitor
corepack enable
pnpm install
pnpm build
pnpm --filter codinfy-agent-monitor link --globalcodinfy-agent-monitor init
codinfy-agent-monitor statusYour private operational data is created inside .codinfy-agent-monitor/ and excluded from Git.
# Animated terminal mission control
codinfy-agent-monitor watch
# Local Fastify + WebSocket dashboard
codinfy-agent-monitor webOpen http://localhost:3579/codinfy. /dashboard and /codinfy-agent-monitor remain supported aliases. The server binds to 127.0.0.1 and rejects foreign Host/Origin values.
Without a global link:
pnpm codinfy status
pnpm codinfy watch
pnpm mcp◫ Command center
╭────────────────── Codinfy Agent Monitor ──────────────────╮
│ Project: codinfy-agent-monitor │
│ Active AI: Codex · Command: /codinfy │
│ MCP: codinfy-agent-monitor │
├────────────────────────────────────────────────────────────┤
│ Context used ██████████████░░░░░░ 68% official │
│ Current rate ████████░░░░░░░░░░░░ 42% estimated │
│ Daily limit ████████████████░░░░ 81% estimated │
│ Weekly limit ████████░░░░░░░░░░░░ 42% estimated │
├────────────── AI Credit Saver & Smart Router ──────────────┤
│ Current model: Premium Reasoning │
│ Recommended: Standard Code · estimated saving: 45% │
│ Model switch: confirmation required │
├────────────────────────────────────────────────────────────┤
│ Agents: 5 active · 3 idle · Workflow: 7 tasks │
│ ✓ secrets.clean ✓ tests.passed ✓ build.passed │
╰────────────────────────────────────────────────────────────╯
© CODINFY PLATFORMS SASU · codinfy.comInterface | Command | Experience |
Snapshot |
| Fast operational overview |
Animated TUI |
| Live terminal bars and agent activity |
Web glass UI |
| Responsive local dashboard + WebSocket |
MCP server |
| Tool API for compatible AI environments |
Slash command |
| Native host-oriented monitoring flow |
⌘ Operator commands
codinfy-agent-monitor status
codinfy-agent-monitor agents
codinfy-agent-monitor context
codinfy-agent-monitor limits
codinfy-agent-monitor workflow
codinfy-agent-monitor tasks
codinfy-agent-monitor timeline
codinfy-agent-monitor files
codinfy-agent-monitor healthcodinfy-agent-monitor git
codinfy-agent-monitor diff --full
codinfy-agent-monitor tests --run
codinfy-agent-monitor build --run
codinfy-agent-monitor observer
codinfy-agent-monitor deps
codinfy-agent-monitor history
codinfy-agent-monitor export --format html
codinfy-agent-monitor secrets
codinfy-agent-monitor check-command "git push --force"
codinfy-agent-monitor attribution-check
codinfy-agent-monitor review
codinfy-agent-monitor public-ready
codinfy-agent-monitor commit-message
codinfy-agent-monitor pr
codinfy-agent-monitor docs-check
codinfy-agent-monitor handoffcodinfy-agent-monitor saver
codinfy-agent-monitor budget
codinfy-agent-monitor cost
codinfy-agent-monitor model-score "audit the authentication architecture"
codinfy-agent-monitor model-advice "update the README"
codinfy-agent-monitor model-rules
codinfy-agent-monitor switch-model standard_code
codinfy-agent-monitor economy-planThe router recommends configurable categories — fast_cheap, standard_code, advanced_code or premium_reasoning — instead of hard-coding provider model names. It never changes the active model without confirmation.
codinfy-agent-monitor beginner
codinfy-agent-monitor next
codinfy-agent-monitor checklist
codinfy-agent-monitor glossary
codinfy-agent-monitor explain-error
codinfy-agent-monitor simple-report
codinfy-agent-monitor install-guide
codinfy-agent-monitor github-guide
codinfy-agent-monitor learncodinfy-agent-monitor language auto
codinfy-agent-monitor language fr
codinfy-agent-monitor level beginner
codinfy-agent-monitor level expert
codinfy-agent-monitor safe on
codinfy-agent-monitor environment
codinfy-agent-monitor doctorcodinfy-agent-monitor node servers
codinfy-agent-monitor node ports
codinfy-agent-monitor node orphans
codinfy-agent-monitor node inspect 24132
codinfy-agent-monitor node cleanup
codinfy-agent-monitor process-map
codinfy-agent-monitor resource-guardStopping a process requires node stop <pid> --confirm. Force Kill additionally requires --graceful-attempted and a second confirmation. Protected or unidentified processes remain inspect-only.
codinfy-agent-monitor update check
codinfy-agent-monitor update changelog
codinfy-agent-monitor update preflight
codinfy-agent-monitor update settings
codinfy-agent-monitor backup
codinfy-agent-monitor backup list
codinfy-agent-monitor doctor
codinfy-agent-monitor recovery
codinfy-agent-monitor notificationsAuto-install is always OFF. update install, update rollback, and backup restore require explicit confirmation and create or validate safety backups.
⬡ MCP Tool API
Start the local stdio server:
codinfy-agent-monitor mcp{
"mcpServers": {
"codinfy-agent-monitor": {
"command": "codinfy-agent-monitor",
"args": ["mcp"]
}
}
}Tool domain | MCP tools |
Status |
|
Agents |
|
Usage |
|
Model economy |
|
Workflow |
|
Engineering |
|
Release |
|
Guided |
|
Trust |
|
Reports |
|
Node runtime |
|
Process safety |
|
Updates |
|
Continuity |
|
The integration test launches the built stdio server, connects an MCP client, lists tools and calls monitor.get_attribution.
⇄ AI environment adapters
Ready-to-copy templates live under templates/. If a host does not expose custom slash commands, the CLI, TUI and local web dashboard remain universal fallbacks.
⎔ Architecture
flowchart LR
subgraph Hosts[AI environments]
Claude[Claude Code]
Codex[Codex]
Cursor[Cursor]
Windsurf[Windsurf]
end
subgraph Interfaces[Codinfy interfaces]
MCP[MCP stdio · 66 tools]
CLI[CLI · /codinfy]
TUI[Animated Ink TUI]
WEB[Fastify + WebSocket]
end
CORE[Core monitoring engine]
DB[(Local SQLite · WAL)]
OPS[Git · Tests · Build · Environment]
TRUST[Secret Scanner · Safe Guard · Review]
ROUTER[AI Credit Saver · Model Router]
Claude & Codex & Cursor & Windsurf --> MCP
MCP & CLI & TUI & WEB --> CORE
CORE --> DB
CORE --> OPS
CORE --> TRUST
CORE --> ROUTERpackages/core domain, SQLite, Git, security, model router, reports
packages/cli codinfy-agent-monitor + codinfy-monitor binaries
packages/tui animated Ink terminal mission control
packages/server local Fastify + WebSocket glass dashboard
packages/mcp-server stdio MCP server codinfy-agent-monitor
packages/adapter-* Claude Code, Codex, Cursor and Windsurf metadata
templates/ ready-to-copy /codinfy and MCP integrationsSee the full architecture guide.
⛨ Trust center
Control | Default |
Data location | Local |
Dashboard exposure | Loopback only — |
Secret output | Redacted before CLI, MCP and report output |
Project commands | Tests/build run only with explicit |
Model switching | Never automatic; confirmation is mandatory |
Git behavior | Read-only monitoring; no implicit commit or push |
Public-ready gate | Secrets + attribution + tests + build + sensitive files |
Never paste API keys, tokens, passwords,.env content, private logs or production credentials into a public issue. Use private vulnerability reporting.
Read the Security Policy · Read the threat model
▣ Requirements
Runtime | Requirement |
Node.js |
|
pnpm |
|
Git | Recommended for file and repository monitoring |
Platforms | Windows, macOS and Linux by design |
◇ Build with us
corepack pnpm install
pnpm checkpnpm check runs lint, 31 tests — including live MCP stdio, host-hook and complete dashboard-route tests — TypeScript build, formatting, secret scanning and attribution validation.
Guide | Purpose |
Local setup, branch flow and PR checklist | |
Private disclosure and support policy | |
Codinfy Attribution License 1.0 | |
Release history | |
Installation, usage, MCP and security |
◉ Local storage
.codinfy-agent-monitor/
├── config.json
├── metrics.sqlite
├── sessions/
├── agents/
├── workflows/
├── logs/
├── reports/
└── cache/The entire directory is ignored by Git. .env*, credentials, logs, build output and local reports are never intended for publication.
⚖ License & attribution
This public repository is distributed under the custom Codinfy Agent Monitor Attribution License 1.0. Because it protects product identity and carries enhanced attribution conditions, it may be classified as source-available rather than OSI open source. Qualified legal review is recommended before high-stakes reliance.
The following identity must remain visible in copies, forks, distributions, interfaces, reports, exports and modified versions:
Codinfy Agent Monitor
/codinfy
codinfy-agent-monitor
© CODINFY PLATFORMS SASU
codinfy.com
Created by CODINFY PLATFORMS SASU
Bakala Goin — Founder & CEOConnect with Codinfy
Bakala Goin — Founder & CEO
Created by CODINFY PLATFORMS SASU Bakala Goin — Founder & CEO codinfy.com
© CODINFY PLATFORMS SASU · codinfy.com
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Latest Blog Posts
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/bakalagoin/codinfy-agent-monitor'
If you have feedback or need assistance with the MCP directory API, please join our Discord server