# glick.run

> glick.run is a portfolio of four self-hosted, agent-native tools you run on your own machines and drive with your AI agent: AppCrane (deploy and run the apps your AI builds), MoorAI (runtime guardrails for AI agents, apps and APIs), Productizer (one living spec per product, kept in your own repo) and IT-AI (self-hosted, AI-powered remote IT support that fixes a machine only after a human approves).

Each tool runs on hardware you own; licences differ by product and are stated on each product's own pages. Productizer and IT-AI are published here on glick.run. AppCrane and MoorAI each have their own site, with its own llms.txt and full-text llms-full.txt: https://appcrane.dev/llms.txt and https://moorai.dev/llms.txt. Old glick.run URLs for AppCrane and MoorAI pages redirect permanently to those sites.

This file lists every page in https://glick.run/sitemap.xml exactly once. A longer reference for Productizer and IT-AI is at https://glick.run/llms-full.txt.

## glick.run

- [glick.run — self-hosted, agent-native AI dev tools](https://glick.run/): The home page: the four tools, what they share (self-hosted, agent-native, open source, private by design) and where each one lives.
- [Blog](https://glick.run/blog/): Index of every post across the products. MoorAI posts are published on moorai.dev and AppCrane posts on appcrane.dev; the index links to them there.
- [OSS Friends](https://glick.run/oss-friends.html): Open-source projects glick.run stands alongside: self-hosted, privacy-first, developer-owned tools, from scheduling and analytics to deployment and secrets.
- [Contact](https://glick.run/contact.html): Questions about MoorAI, AppCrane, Productizer or IT-AI, pricing, or a deployment.
- [Log in](https://glick.run/login.html): Log in to the hosted MoorAI console with a one-time email link.
- [Start free](https://glick.run/signup.html): Create a free MoorAI console account: policy, fleet visibility and compliance evidence for the on-device agent.

## Productizer — one living spec per product, in your own repo

> Productizer is an open-source Claude Code plugin that keeps a single living spec for your product inside your own repository. Requirements are written in EARS form with permanent ids that are never reused or renumbered. Every new intent is classified against the whole spec (extend, refine, duplicate or contradict), and a contradiction halts the work until a human rules on it. A constitution of P-numbered principles sits above the requirements and is checked first. Runs locally: no account, no server, no telemetry. Source: https://github.com/gitayg/productizer

- [Productizer](https://glick.run/productizer.html): Productizer is an open-source Claude Code plugin that keeps one living spec per product. Every intent is checked against it; a contradiction stops the work.
- [The ten things Productizer does](https://glick.run/productizer-features.html): One living spec, permanent EARS ids, every intent classified, a contradiction halt, human gates, a read-only dashboard, and what is not yet proven.
- [Productizer vs the field](https://glick.run/productizer-comparison.html): Productizer vs AI Unified Process, Entire, 8090, Lightsprint, Gas City and Credo AI: ten structural properties, and where each rival is ahead.
- [Productizer compliance coverage](https://glick.run/productizer-compliance.html): What Productizer's artifacts evidence across NIST AI RMF, NIST SSDF, the EU AI Act, SOC 2 and ISO 42001, and the rows it covers nothing for.
- [Regulating AI-written code — what changed in 2026](https://glick.run/productizer-regulation.html): OMB M-26-05 rescinded the US attestation mandate, EU AI Act high-risk dates slipped, and NIST does not single out AI-generated code. Sources quoted.
- [Productizer vs AI Unified Process](https://glick.run/productizer-vs-aiup.html): One living spec that halts on contradiction vs a use-case-per-file method with a working CI lifecycle, including where AIUP wins.
- [Productizer vs Entire](https://glick.run/productizer-vs-entire.html): Entire captures and redacts agent sessions and scores diffs with in-repo gates; Productizer declares one living spec up front and halts on contradiction.
- [Productizer vs 8090 Software Factory](https://glick.run/productizer-vs-8090.html): A hosted AI-native SDLC platform for regulated enterprises vs a free, open-source Claude Code plugin.
- [Productizer vs Lightsprint](https://glick.run/productizer-vs-lightsprint.html): Lightsprint scores PR readiness from observed GitHub evidence; Productizer decides what should be true before work starts.
- [Productizer vs Gas City](https://glick.run/productizer-vs-gascity.html): Gas City orchestrates many coding agents over a Dolt-backed work graph; Productizer declares one living spec and halts on contradiction.
- [Productizer vs Credo AI](https://glick.run/productizer-vs-credoai.html): Both hook into Claude Code the same way; Credo AI Agent Governor governs the agent's runtime behaviour, Productizer governs the agreement.

## IT-AI — the self-hosted RMM your AI can operate

> IT-AI is a self-hosted RMM (remote monitoring and management) for Windows, macOS and Linux, built to be operated by your AI. You run the hub (Elastic License 2.0, free for your own organization); a lightweight agent on each device (MIT) dials out to it, so no inbound ports are needed. An MCP server with 32 tools and the itai CLI (both MIT) let Claude Code or another MCP client drive the fleet, bounded by scoped tokens, a hub deny-list and an audit log. The hub's own assistant applies a fix only after a person clicks Apply. Every page below also has a plain-Markdown twin at the URL shown after it.

- [IT-AI](https://glick.run/it-ai/): The self-hosted RMM your AI can operate safely: what the AI can and cannot do without you, capabilities, architecture, FAQ. Markdown: https://glick.run/it-ai.md
- [How IT-AI works](https://glick.run/it-ai/how-it-works.html): Outbound-only relay, per-device secrets, token scopes, deny-list, audit, LAN-direct capabilities and signed self-update. Markdown: https://glick.run/it-ai/how-it-works.md
- [IT-AI MCP server and itai CLI](https://glick.run/it-ai/mcp.html): All 32 MCP tools, which ones a read-only token can call, every itai command, and setup for Claude Code and other MCP clients. Markdown: https://glick.run/it-ai/mcp.md
- [IT-AI security and trust](https://glick.run/it-ai/security.html): Buyer questions answered, including the gaps: no OS code signing yet, in-memory audit log, no certifications. Markdown: https://glick.run/it-ai/security.md
- [IT-AI pricing](https://glick.run/it-ai/pricing.html): Free to self-host (MIT agent, Elastic License 2.0 hub); hosted plan coming, no prices published. Markdown: https://glick.run/it-ai/pricing.md
- [IT-AI documentation](https://glick.run/it-ai/docs.html): Run the hub, enroll a device, connect your AI, background jobs, hardening settings. Markdown: https://glick.run/it-ai/docs.md
- [IT-AI compared](https://glick.run/it-ai/compare.html): Feature table against NinjaOne, Atera, Action1, Level, Tactical RMM, MeshCentral, Breeze, TeamViewer and RustDesk, sourced and dated. Markdown: https://glick.run/it-ai/compare.md
- [IT-AI use cases](https://glick.run/it-ai/use-cases.html): Internal IT, small MSPs, and edge and robotics fleets behind CGNAT, and when IT-AI is the wrong tool. Markdown: https://glick.run/it-ai/use-cases.md
- [Download IT-AI](https://glick.run/it-ai/download.html): Agent, MCP server and CLI files per platform, and how to verify them. Markdown: https://glick.run/it-ai/download.md
- [IT-AI agent changelog](https://glick.run/it-ai/changelog.html): Every agent release since 3.3, newest first. Markdown: https://glick.run/it-ai/changelog.md
- [IT-AI vs TeamViewer](https://glick.run/it-ai-vs-teamviewer.html): A self-hosted AI RMM vs the cloud remote-access incumbent: your own hub, no per-seat fees, fixes only with approval.
- [IT-AI vs RustDesk](https://glick.run/it-ai-vs-rustdesk.html): RustDesk is a mature self-hosted 1:1 remote desktop; IT-AI is an AI RMM with fleet health and an MCP server.
- [IT-AI vs NinjaOne](https://glick.run/it-ai-vs-ninjaone.html): NinjaOne is a full-suite cloud RMM with patching and ticketing at scale; IT-AI is a lighter, self-hosted AI RMM your AI can operate.

## AppCrane — on appcrane.dev

> AppCrane is the self-hosted, agent-native Heroku and Vercel alternative: deploy and run the apps your AI builds on a server you own, driven by your coding agent over MCP, with sandbox and production per app, Docker isolation, auto-HTTPS, SSO and a per-user audit trail. Every AppCrane page, including its comparisons, is listed in https://appcrane.dev/llms.txt.

- [AppCrane](https://appcrane.dev/): The product site: self-hosted deploys driven by your coding agent over MCP, to a server you own.
- [Deploy from Claude & Cursor](https://appcrane.dev/deploy-from-claude.html): Connect your coding agent over MCP and let it deploy to a server you own.
- [AppCrane vs Coolify, Dokploy & Vercel](https://appcrane.dev/comparison.html): The comparison hub, with a page per competitor.
- [Changelog](https://appcrane.dev/changelog.html): Every AppCrane release and what shipped.
- [AppCrane on GitHub](https://github.com/gitayg/appCrane): Source code.

## MoorAI — on moorai.dev

> MoorAI: AI security at the point. On-device AI detection and response, checking every agent action before it runs. Runtime guardrails for AI agents, apps and APIs: coding agents, Agent SDK services, any agent framework through a localhost sidecar, remote MCP servers through an MCP gateway, and model API calls through a model proxy. MoorAI is open source (MIT). It checks what developers send to Claude Code, Codex, Copilot CLI, Gemini CLI and Cursor, and what those agents read, run and reply, on the developer's machine (enforcement is validated end to end on Claude Code; the other four run the same checks through their own pre-tool hooks); security teams receive content-free signals (a category, a risk level and a keyed one-way hash, plus the login and hostname, which the console pseudonymises on arrival), never the prompt. The agent ships for macOS (Apple silicon) and Windows (x64). Every MoorAI page, including the per-competitor comparisons, is listed in https://moorai.dev/llms.txt.

- [MoorAI](https://moorai.dev/): The product site: runtime guardrails for AI agents, apps and APIs, and the optional console that receives content-free signals.
- [MoorAI vs the field](https://moorai.dev/moorai-comparison.html): The comparison hub, with a page per competitor.
- [MITRE ATLAS agentic-technique coverage by vendor](https://moorai.dev/moorai-atlas-coverage.html): What 43 AI-security vendors document against the 76 MITRE ATLAS agentic techniques, every credit with a quote and a URL, MoorAI scored the same way.
- [MITRE ATT&CK v20 coverage, technique by technique](https://moorai.dev/moorai-attack-coverage.html): Every MoorAI rule mapped to MITRE ATT&CK Enterprise v20, each in-scope technique with the rules that credit it, the coverage level, the reason and the lane that can see it.
- [Agentic security benchmark](https://moorai.dev/benchmark.html): The public corpora and harness MoorAI is scored against, and how to score your own product.
- [For developers using AI coding agents](https://moorai.dev/for-ai-coding-agents.html): Hooks in Claude Code, Codex, Copilot CLI, Gemini CLI and Cursor that check each shell command, file read and MCP call on the device.
- [MoorAI for Agents](https://moorai.dev/moorai-for-endpoints.html): The product for the devices people work on: coding-agent hooks, the desktop app, a browser extension for web AI tools, on-device DLP, shadow-AI detection and the local AI inventory.
- [MoorAI for AI Servers](https://moorai.dev/for-ai-apps.html): The product for the AI you ship: Agent SDK hooks, a sidecar, a model proxy, an MCP gateway and an egress proxy for AI apps, agents, RAG pipelines and APIs.
- [MoorAI console](https://moorai.dev/console.html): Every agent, MCP server, model and device on the fleet, in one view, built from signals, never prompts or code.
- [How MoorAI works](https://moorai.dev/how-it-works.html): One check where the AI acts, before it acts, and content-free signals to the console.
- [AI agents in digital forensics and incident response](https://moorai.dev/ai-in-digital-forensics.html): For DFIR teams using AI agents on case data: the GenAI-assisted weaknesses in SOLVE-IT, what MoorAI mitigates (partial checks labelled, alert basis, evidence-path and collection-host blocks, a tamper-evident action record) and what it does not.
- [Download MoorAI](https://moorai.dev/download.html): The macOS and Windows agents.
- [Quick start](https://moorai.dev/quick-start.html): Install the Claude Code hooks with one command, add other agents, check the install, and enroll to enforce.
- [Claude Code security](https://moorai.dev/claude-code-security.html): How to stop Claude Code reading .env or leaking secrets: native deny rules plus an on-device PreToolUse hook.
- [MCP security](https://moorai.dev/mcp-security.html): How to secure the MCP servers coding agents use: approval, argument and result scanning, tool poisoning, rug pulls.
- [MoorAI on GitHub](https://github.com/gitayg/moorai): Source code.

## Blog posts

Every post in the glick.run blog index, newest first. Each is published on its product's site.

- [When the console is down: what to ask any AI-agent security vendor](https://moorai.dev/blog/ai-agent-security-when-the-console-is-down.html): 2026-10-22, MoorAI
- [The future of application platforms defends itself.](https://appcrane.dev/blog/appcrane-not-the-last-line-of-defence.html): 2026-10-22, AppCrane
- [When a coding agent weakens oversight: hold it, disclose it, page on it](https://moorai.dev/blog/ai-coding-agents-weakening-oversight.html): 2026-10-22, MoorAI
- [Local LLMs and AI agents in air-gapped OT: the gap stops the cloud, not the agent](https://moorai.dev/blog/local-llms-and-ai-agents-in-air-gapped-ot.html): 2026-10-21, MoorAI
- [Mac Studio as a local AI server: what to lock down](https://moorai.dev/blog/protect-mac-studio-local-ai.html): 2026-10-21, MoorAI
- [Damn Vulnerable MCP: MoorAI blocks 9 of 10 outright, 10 of 10 with IP exposure set to block](https://moorai.dev/blog/damn-vulnerable-mcp-moorai-results.html): 2026-10-17, MoorAI
- [ISO/IEC 27090 names the threats. MoorAI stops them where AI agents act.](https://moorai.dev/blog/iso-27090-for-ai-agents-on-the-device.html): 2026-10-17, MoorAI
- [AEGIS for AI agents on the device: what each domain asks and what you can measure](https://moorai.dev/blog/aegis-for-ai-agents-on-the-device.html): 2026-10-16, MoorAI
- [MoorAI and Microsoft Execution Containers (MXC)](https://moorai.dev/blog/moorai-and-microsoft-execution-containers-mxc.html): 2026-10-14, MoorAI
- [Backdoored models show up in what they do](https://moorai.dev/blog/backdoored-models-show-up-in-what-they-do.html): 2026-10-08, MoorAI
- [Hidden Unicode in what AI agents read](https://moorai.dev/blog/hidden-unicode-in-what-ai-agents-read.html): 2026-10-07, MoorAI
- [How MoorAI protects AI apps and APIs](https://moorai.dev/blog/how-moorai-protects-ai-apps-and-apis.html): 2026-10-06, MoorAI
- [MoorAI against the Synthetic Insider Threat Matrix™](https://moorai.dev/blog/moorai-against-synthetic-insider-threat-matrix.html): 2026-10-01, MoorAI
- [Latio's second era of AI security: where MoorAI fits](https://moorai.dev/blog/latio-second-era-ai-security.html): 2026-09-17, MoorAI
- [Every boundary an AI coding agent crosses — mapped to MITRE ATLAS](https://moorai.dev/blog/agent-boundaries-mapped-to-mitre-atlas.html): 2026-09-16, MoorAI
- [The layer above the model — ASD says security lives in the harness](https://moorai.dev/blog/the-layer-above-the-model.html): 2026-09-14, MoorAI
- [Agentic apps are attack graphs — and the guardrail is the missing node](https://moorai.dev/blog/agentic-apps-are-attack-graphs.html): 2026-09-13, MoorAI
- [Sixty-seven open-source business apps, two clicks each](https://appcrane.dev/blog/appcrane-app-catalogue.html): 2026-09-09, AppCrane
- [How to choose your agentic security vendor](https://moorai.dev/blog/choosing-an-agentic-security-vendor.html): 2026-09-06, MoorAI
- [MoorAI, graded against AMTSO's agentic testing guidelines](https://moorai.dev/blog/amtso-agentic-security-testing-guidelines.html): 2026-09-06, MoorAI
- [Obfuscated prompts beat content scanning. Not the action layer.](https://moorai.dev/blog/obfuscated-prompts-beat-content-scanning.html): 2026-09-03, MoorAI
- [Your AI writes the code. Who attests to it?](https://moorai.dev/blog/ai-supply-chain-attestation-content-free.html): 2026-09-02, MoorAI
- [AI's Evidence Problem — answered content-free](https://moorai.dev/blog/ai-evidence-problem-content-free.html): 2026-08-28, MoorAI
- [Prompt injection: content scanning won't stop it](https://moorai.dev/blog/prompt-injection-ai-assistant-attacker.html): 2026-08-28, MoorAI
- [76% own AI governance. 54% have no way to audit it.](https://moorai.dev/blog/moorai-sans-2026.html): 2026-08-22, MoorAI
- [MoorAI & ISO/IEC 42001](https://moorai.dev/blog/moorai-iso-42001.html): 2026-08-20, MoorAI
- [The 5 questions every AI-governance buyer is being asked](https://moorai.dev/blog/five-questions-ai-governance.html): 2026-08-18, MoorAI
- [AI Bill of Materials (AIBOM) & the EU AI Act](https://moorai.dev/blog/ai-bill-of-materials-eu-ai-act.html): 2026-07-21, MoorAI
- [Why AppCrane now speaks MCP](https://appcrane.dev/blog/why-appcrane-speaks-mcp.html): 2026-06-28, AppCrane

## Contact

- [GitHub](https://github.com/gitayg): All four projects' source repositories.