# Share Law Firm Skills

Use the [project website](https://lawfirmskills.com/) as the
single introduction link. Use the [repository](https://github.com/sammollaei/law-firm-skills)
for source, installation, issues, and attribution. A directory listing is not
live until its acceptance and published URL have been verified.

## Message for a friend

I built 23 free AI workflows for lawyers and law firm owners. They help with
morning briefs, meeting follow-ups, decisions, and staff playbooks. Start with
one workflow and a fictional example. Setup and source:
https://lawfirmskills.com/

If you use Claude Code, run these inside Claude Code:

```text
/plugin marketplace add sammollaei/law-firm-skills
/plugin install law-firm-skills@sam-mollaei
```

Restart, then try `/law-firm-skills:setup`. For Codex and individual skills, use
the [installation guide](installation.md). The collection does not connect
your accounts; you choose what notes or authorized tools it can use.

## Public post

Lawyers do not need another pile of prompts they never open.

I built Law Firm Skills: 23 repeatable AI workflows for the actual week of a
law firm owner. A morning brief. A meeting turned into accountable next steps.
A process your team can follow without asking you the same questions again.

The workflows distinguish completed work from claims, drafts from sends, and
routine follow-through from decisions that need you. Start with a fictional
example before supplying your own information.

Free, open source, and published under my name, Sam Mollaei:
https://lawfirmskills.com/

Try one and tell me where the instructions could be clearer.

## Three one-minute walkthroughs

These silent, captioned videos present excerpts of actual Claude Code test
outputs from fictional inputs. They are designed transcript walkthroughs,
not screen recordings or demonstrations of connected accounts. They do not
claim savings, real client results, or guaranteed model behavior.

- [Morning brief](https://github.com/sammollaei/law-firm-skills/releases/download/v0.2.0/law-firm-skills-morning-demo.mp4)
- [Choosing a first move](https://github.com/sammollaei/law-firm-skills/releases/download/v0.2.0/law-firm-skills-decide-demo.mp4)
- [Meeting decisions and boundaries](https://github.com/sammollaei/law-firm-skills/releases/download/v0.2.0/law-firm-skills-meeting-demo.mp4)

Each video is 60 seconds at 1920x1080. The scripts below are also available for
Sam to record a live host walkthrough in his own voice.

## Three demonstration scripts

Record only fictional material. Show the real host loading the intended public
skill and the actual answer. Do not substitute a designed mockup for a tested
workflow or present these scripts as recorded demonstrations.

### 1. A useful first decision (60-90 seconds)

- 0-15 seconds: open the install guide and choose decide. Explain that no inbox
  connection is needed for this example.
- 15-35 seconds: invoke decide with the fictional reminder-versus-CRM scenario
  from the homepage. Show the supplied facts and the advice-only boundary.
- 35-65 seconds: show the actual recommendation and the evidence that would
  change it. Point out that no savings or completed setup were invented.
- 65-90 seconds: show the source link and invite viewers to try that example.

### 2. A morning brief you can trust (60-90 seconds)

- 0-15 seconds: select morning and show the fictional snapshot in the install
  guide. State that live systems are not connected for this demonstration.
- 15-40 seconds: run the brief. Highlight the unsent client reply and the SOP
  that is claimed complete but lacks its document.
- 40-65 seconds: show coverage gaps and how routine evidence stays with the
  existing owner. Explain the distinction between draft and completed work.
- 65-90 seconds: point to configuration for the viewer's own sources and rules.

### 3. A meeting becomes accountable work (60-90 seconds)

- 0-20 seconds: supply fictional notes: Alex will draft a checklist by October
  20; Morgan will review once it exists; CRM replacement was discussed only.
- 20-45 seconds: invoke meeting, requesting proposals only. Show the resulting
  owners, dependency, stated date, and unresolved decision.
- 45-70 seconds: show that discussion did not become approval and no message or
  task was sent or created. Explain what a separately approved action would need.
- 70-90 seconds: show the collection link and ask viewers which handoff is hardest.

## Pilot feedback template

Invite willing participants directly; publishing this file is not an invitation
sent to anyone. Use their chosen private channel for feedback. No client or
employee details are needed.

1. Platform/version and skill tried.
2. Minutes from opening the guide to first useful result.
3. What worked, what was confusing, and whether any claim lacked evidence.
4. Whether they used it again in seven days and which workflow.
5. Whether an anonymized example may be quoted publicly (separate permission).

The initial pilot targets are 8/10 installs within ten minutes and 6/10 useful
first results. Seek four return users in the next week before expanding outreach.
These are test targets, not achieved results or predicted adoption. Measure
firms separately from individual users. Stars and installs are not active usage.

## Distribution order

1. Verify the public GitHub install and website; share one clear example.
2. Use the skills CLI's documented discovery process and verify actual search
   results. Do not generate artificial installs or stars.
3. Submit the skills-only plugin through the verified OpenAI publisher account
   and Anthropic directory. Publish only accepted, tested listings.
4. Test OpenClaw before publishing skills to ClawHub under the actual publisher
   namespace; a successful upload is not a passed security scan.
5. Share through Sam's own audience and opt-in pilots. Contribute to relevant
   lawyer communities according to their posting rules. No bulk unsolicited DMs.

Account verification, directory approval, pilot participation, case-study
permission, and repeat use are external steps. Do not describe them as complete
on the strength of prepared files or submitted forms.
