real-estate-pack

Compliance guardrails, skills, loops & review agents for US residential real estate agents on Claude Code.

Most "AI for real estate" resources are copy-paste prompt packs. This one is enforced: the guarantees come from hooks (deterministic - they run every time), with rules, skills, loops, and review agents on top. It is built around the two things that get agents in trouble with AI: Fair Housing and fabricated facts.

Every item serves at least one goal:

Icon Goal
🪙 Fewer tokens
🎯 Accuracy (verified facts)
✅ Truth-telling (no hallucinated property details)
⚡ Speed
💎 Quality / compliance standard

⚠️ Important - read first

  • This is not legal advice. It is a drafting + guardrail tool. You, the licensed agent, are responsible for everything you publish or send.
  • US residential focus. Fair Housing (federal), NAR, and disclosure references are US-based. Outside the US, swap in your jurisdiction's law.
  • Hooks are guardrails, not guarantees. fair-housing-guard catches common, documented phrases - not every possible violation. Always run the compliance-reviewer and read every word before publishing.
  • State and brokerage rules vary (disclosures especially). Confirm with your broker/attorney.

What's inside

Folder Contents Layer
rules/ Advisory CLAUDE.md fragments (Fair Housing, truth-telling, confidentiality, no-advice, disclosure) Advisory
skills/ Listing, CMA, consults, offers, market reports, lead follow-up, etc. On-demand
hooks/ Deterministic enforcement (fair-housing, PII, fact-check, superlative, photo disclosure, review) + settings.json Enforced
loops/ Multi-step slash commands (/list-it, /price-it, /offer-review, ...) Workflow
agents/ Review subagents (compliance / listing / contract) Review
mcp/ Browser MCP for verifying public property/market facts Tooling
templates/ Listing intake, CMA, transaction checklist, client brief Docs
tests/ test_hooks.py - runs every hook against sample inputs Proof

See INVENTORY.md for the full table of every item and why it exists.

Install

Windows (PowerShell):

./install.ps1 -ProjectPath "C:\path\to\your\workspace"

macOS / Linux:

./install.sh /path/to/your/workspace

The installer copies hooks/ into <project>/.claude/hooks/, merges hooks/settings.json into <project>/.claude/settings.json, appends rules/ into <project>/CLAUDE.md, and copies skills -> .claude/skills/, loops -> .claude/commands/, agents -> .claude/agents/. (mcp/ and templates/ are for reference - wire them in per the folder READMEs.)

Requires: Python 3.8+ on PATH (the hooks are Python; they run on Windows/macOS/Linux).

Verify the pack itself: python tests/test_hooks.py (16 cases across all 7 hooks).

Uninstall (removes only what it installed, restores your CLAUDE.md):

./uninstall.ps1 -ProjectPath "C:\path\to\your\workspace"   # Windows
./uninstall.sh /path/to/your/workspace                      # macOS / Linux

Slash commands

After install, type / in Claude Code to see them:

  • Workflows (multi-step): /list-it, /price-it, /new-lead, /offer-review, /close-it, /market-update.
  • Skills (one slash command per skill): /cma, /listing-description, /buyer-consultation, /seller-consultation, /offer-analysis, /market-report, /lead-followup, /open-house-plan, /transaction-timeline, /disclosure-checklist, /social-content, /objection-handling, /negotiation-strategy, /farming-prospecting, /truthful-reporter.
  • Review agents (one per agent): /compliance-reviewer, /listing-reviewer, /contract-reviewer.

Skills and agents also trigger automatically - the slash command is just the explicit way to call one.

Honest scope

  • New repo - starts at 0 stars. No inflated claims.
  • Inspired by the dev-pack methodology (github.com/m-binimran/dev-pack) and grounded in published Fair Housing advertising guidance and 2026 real-estate-AI compliance writing.
  • The guardrails reduce risk; they do not eliminate it. Compliance is the licensed agent's responsibility.

License

MIT - see LICENSE.