Update (June 2026): We have since open-sourced cli-bridge, a small, local, open-source tool that lets Hermes and other bots run on a terminal AI CLI you already pay for. Read the new article, We've Open-Sourced a CLI Bridge for Your Claude Code Subscription, or go straight to the GitHub repository. Please read the disclaimer and risk notice before you use it.
How Can I Use OpenClaw? 10 Use Cases From Our Firm and the Legal Issues You Should Know About
AI agents are everywhere in legal tech marketing right now. Harvey, Clio, Spellbook - the pitch is always the same: hand over your data, pay per seat, and trust the cloud. For a law firm handling sensitive client matters across multiple jurisdictions, that pitch has always sat uncomfortably for us here at Daimon Legal.
So we built something different. Over the past several months, we have been running AI agents on self-hosted infrastructure in our own homelab - a single VM drawing minimal power, connected exclusively via Tailscale, with segregated network, file system and resources. Our workflows, memory and files stay on our infrastructure. No third-party SaaS controls our agent architecture.
Here are ten practical use cases that have changed how this firm operates.
TLDR:
- AI agents can handle far more than document review and legal research
- Self-hosting gives law firms control over agent architecture, but LLM API data flows need careful management
- Ten working use cases covering SEO, legislative monitoring, CRM, travel, research and more
- You need patience and some technical willingness, but the payoff is real
1. Website Analytics and SEO Monitoring
The agent hooks into Google Search Console and DataForSEO to run daily analytics on the firm's website. It identifies keyword opportunities, flags website health issues and tracks traffic patterns. Instead of paying for an SEO agency retainer or logging into three different dashboards each morning, the daily report lands in one place with actionable items already prioritised.
2. Daily News Digest and Industry Monitoring
A daily digest pulls from Hacker News, X, Reddit and blog subscriptions, filtered for relevance to our practice areas. The agent runs a vibe-check across sources and delivers a curated summary. Think of it as a custom Bloomberg terminal for legal tech and regulatory developments, sent to its own dedicated channel each morning.
3. Self-Hosted RSS Feed Management
The agent spins up custom RSS feeds for our self-hosted Omnivore reader. New feeds take minutes to create, even when the source site or newsletter is not RSS-friendly. This solves a persistent annoyance: valuable legal commentary locked behind email newsletters or non-standard publishing formats, now normalised into a single reading workflow.
4. Conference and Event Planning
Feed the agent a conference schedule and it runs a structured event planning process starting four weeks out, with increasing frequency as the event approaches. It does deep dives on speakers, identifies relevant side events and talks, builds calendar entries and manages logistics end-to-end. For a firm that attends international conferences regularly, this replaced hours of manual preparation per event.
5. Restaurant and Cafe Recommendations on the Road
When travelling for conferences or client meetings, the agent provides restaurant and cafe suggestions based on Google Reviews and foodie blogs, complete with Google Maps routing and direct links. Not exactly rocket science on its own, but the agent already knows dietary preferences, budget ranges and the kind of places worth walking an extra block for.
6. Holiday and Travel Planning
The agent creates full holiday plans with deep research into places to visit. Because it has saved preferences from previous trips, updates and new plans reflect what actually matters rather than generic tourist recommendations. An agent that remembers what you liked last time and adjusts accordingly turns out to be surprisingly useful.
7. Legislative Monitoring and Regulatory Alerts
This is where it gets genuinely valuable for legal practice. The agent pulls legislative updates from official sources and runs a summary of the latest developments in bills and regulatory changes. Set it on a cron job and you get daily alerts when something relevant appears. For a firm advising on AI governance and regulatory strategy, this is not a nice-to-have - it's an essential service.
8. Research Verification and Fact-Checking
The agent runs what we call a BS detector over claims we encounter online. It knows our analytical preferences and writing style, so the output is calibrated rather than generic. In a profession where citing unreliable sources can be career-ending, having automated verification that matches your own rigour is significant.
9. Academic Research and Knowledge Management
The agent finds journal articles and academic research, converts files and websites into readable markdown, then posts to a self-hosted Obsidian vault and Outline knowledge base. The shift from manually bookmarking PDFs to having research automatically processed and filed in your preferred format saves hours of effort. Open source tools make this possible without handing your research library to yet another SaaS provider.
10. CRM Management With On-the-Go Capabilities
The agent manages a self-hosted Twenty CRM with capabilities that are not possible through the CRM's native interface alone. It updates diary entries, calendar items and contact lists. It is like having a legal secretary who also happens to hold a PhD in software engineering, DevOps and operational security - and can build custom solutions on the fly.
Why Self-Hosted Matters for Law Firms
The legal profession has specific obligations in relation to privacy and client confidentiality that make cloud-based AI tools genuinely risky and leave clients exposed to potential loss of legal professional privilege if not managed carefully. When client data passes through a third-party AI platform, the question of who controls that data - and where it physically resides - becomes a potential compliance problem.
Self-hosting addresses this directly. The entire setup runs on a VM in the homelab with segregated resources. Network access is via Tailscale only - no ports open to the public internet. The infrastructure draws minimal power and runs 24/7 without requiring a laptop to stay on.
However there is some trade-off: you either need access to a strong technical support team or be prepared to do initial groundwork and problem-solving. However, these are manageable for anyone with the motivation, and the community around open source agentic tools has grown into a genuinely helpful resource for troubleshooting and sharing approaches.
What You Need to Know: Self-hosting your agent infrastructure is only half the picture. Unless you are running local LLMs (such as Ollama or vLLM on your own hardware), every prompt you send to an LLM API - OpenAI, Anthropic, Google - means client data is hitting third-party servers.
API providers generally commit not to train on API data, and enterprise plans may offer zero data retention. However, "generally" and "may" are not words that sit well with Rule 9 of the Australian Solicitors' Conduct Rules, which imposes strict obligations on client confidentiality. And for other businesses, you still have to contend with the Privacy Act which places clear restrictions on how a customer's personal information can be used.
Inputting client-privileged or personal information into any third-party AI tool risks waiver of legal professional privilege not to mention the potential privacy breaches. Businesses need to think hard about what data enters the context window and whether their API agreements, data processing addenda, and retention policies actually hold up under their professional obligations.
The self-hosting approach has many advantages: you control the orchestration layer, the memory, the file system, and the network perimeter. You choose which LLM provider sees which data. That is a major improvement over SaaS tools where the vendor controls the entire stack.
Don't Skip the Governance
Businesses adopting AI agents - whether cloud-based or self-hosted - should be thinking about AI audit and governance from day one. That means documenting what the agents do, what data they access, how outputs are verified, and who is accountable when something goes wrong. Businesses that treat AI adoption as a pure technology project without building governance around it are setting themselves up for problems.
The Final Report
We have been running this setup for several months and it has changed how the firm operates. The SaaS legal AI market is full of tools that promise to revolutionise your practice for $X per seat per month. Some of them are good. All of them require you to trust someone else with your business' sensitive data.
For businesses who care about data sovereignty, who want to understand what their tools are actually doing, and who are willing to put in the initial time - self-hosted AI agents can provide real benefits. They work around the clock. And they get better the longer you use them.
For a confidential discussion about AI governance, compliance, or deploying AI agents in your business, contact Daimon Legal.
The information on this page is general in nature and does not constitute legal advice. Please review our Legal Disclaimer for important information about the limitations of this content and the terms governing your use of this website.