pdombkins turns rose_lawyer into a legal-AI operating suite

This is a broad feature drop that moves the fork well beyond a single legal research or drafting tool.

workflowsecurity

The update adds a menu of capabilities rather than one neat product change:

AI agents that plan work, seek approval, and run linked tasks.
In-app and email notifications.
An audit trail and role-based access controls, including conflict walls.
A selectable AI-model catalogue, including Kimi K3.
Verification through an AI check or outbound research links.
Curated Australian and New Zealand regulatory updates.
Structured tables with reference material, manual edits, and AI questions.
A clause library and playbook-building tools.
Organisation and personal context for AI work.
Administrative usage analytics.
Word, PDF and Markdown exports, including Australian legal citation styling.
Integrations for external AI tools and personal access tokens.
Natural-language workflows and competitor-scanning tools.

Given the scale, this looks best treated as a source of individual ideas, not a drop-in upgrade.

So what Legal-product teams looking for stronger governance, automation and practice-specific workflow ideas should inspect the fork on GitHub.

Spotted something wrong? Or know the PR text has fresher detail than the writeup above?

Commits in this thread

1 commit from pdombkins/rose_lawyer, oldest first. Source extracted verbatim from the harvested git log.

SHA	Subject	Author	Date
`77bc7db6`	feat: 19 competitor features (C002-C040) + Kimi K3 provider	Peter Dombkins	2026-07-20	↗ GitHub
commit body - Agent runtime with plan/approval/DAG executor + /agents page - Notifications (in-app + optional Resend email), audit log + RBAC (project roles, ethical wall) - Data-driven model registry; Moonshot/Kimi K3 via OpenAI-compat (self-host preferred) - Deep-verify (C024): Jade AI check or human self-validation via outbound search links - Regwatch (C018): curated official AU/NZ regulatory RSS feeds - Tabular v2: typed columns, per-column reference docs, manual cell edits, tabular ask - Clause library, playbook builder tools, org/personal context injection - Admin analytics + cohorts, exports (DOCX/PDF/MD + AGLC4 restyle), MCP server + PATs - Workflow NL compile/run; competitor-scan tooling; 9 new migrations

SHA

Subject

Author

Date

77bc7db6

feat: 19 competitor features (C002-C040) + Kimi K3 provider

Peter Dombkins

2026-07-20

↗ GitHub

commit body

- Agent runtime with plan/approval/DAG executor + /agents page
- Notifications (in-app + optional Resend email), audit log + RBAC (project roles, ethical wall)
- Data-driven model registry; Moonshot/Kimi K3 via OpenAI-compat (self-host preferred)
- Deep-verify (C024): Jade AI check or human self-validation via outbound search links
- Regwatch (C018): curated official AU/NZ regulatory RSS feeds
- Tabular v2: typed columns, per-column reference docs, manual cell edits, tabular ask
- Clause library, playbook builder tools, org/personal context injection
- Admin analytics + cohorts, exports (DOCX/PDF/MD + AGLC4 restyle), MCP server + PATs
- Workflow NL compile/run; competitor-scan tooling; 9 new migrations

Capture this thread into my fork

Download a single Markdown prompt that tells Claude how to port every commit above into your working tree — adapting paths and structure to match your repo. Run it via claude -p < capture-thread-934.md from inside the repo you want the changes in.