CRE Agent Skills
github.com/ahacker-1/cre-agent-skills- Category
- AI Agents
- Rank
- No. 502Tools index
- Pricing
- Open Source
- Platform
- cli
- Type
- TOOL
- Builder
- ahacker-1
- GitHub
- 85 stars
- Latest release
- v1.2.0
- Added
- Jul 4, 2026
About
A library of 66 standalone AI skill files for commercial real estate professionals covering underwriting, lease abstraction, investor reporting, capital markets, asset management, and more. Each skill works independently in Claude, ChatGPT, or any LLM tool — no orchestrator, no API keys, no installation required. Built by an AI consulting firm that specializes in deploying AI workflows for CRE operators.
What it does
CRE Agent Skills supplies task-specific prompt playbooks that turn property documents and deal inputs into structured analysis. Users select a playbook, provide its requested data, and optionally add reference material for formulas, benchmarks, or criteria. The collection spans acquisition work, operations, brokerage, office, industrial, and debt recapitalization workflows.
Why it's ranked here
The collection is unusually practical because its prompts map to recognizable CRE deliverables, from rent-roll checks through committee memos and lender packages. New sector packs require companion research, traceable defaults, and explicit geographic assumptions. Still, the repository describes educational decision support, not production software or professional advice, so human verification remains essential.
What's good
Skills specify required inputs, structured outputs, confidence levels, and missing-data flags. Recommended combinations show how to assemble focused workflows without forcing every user into one pipeline. The research standard requires cited support for new benchmarks and formulas, favors primary sources, and asks authors to document conflicting evidence, scope limits, and red flags.
Tradeoffs
Quality depends on the chosen language model and the completeness of supplied documents. Some newer packs are explicitly U.S.-only, while legacy shared material is not always fully sector-neutral. Market-sensitive assumptions still require local validation. The project also avoids orchestration, so users must manage context, sequencing, source documents, and handoffs themselves.
How to use it well
It best suits CRE analysts, owners, brokers, lenders, and asset managers who already have a defined task and source data. Start with one skill, add its recommended references, then chain related skills for diligence, underwriting, reporting, or recap work. Use structured tables where possible and investigate every low-confidence result. It does not replace legal, tax, accounting, financing, or investment professionals.
Technical notes+
README.md describes a Markdown-first repository with 66 skills, 23 knowledge bases, 11 Claude Code plugins, zero dependencies, and three loading patterns: copy one skill into an LLM, install a department plugin for Claude Code, or point an agent at the repository. docs/HOW-TO-USE.md documents Claude Projects, Claude Code, ChatGPT, Cursor, Windsurf, and generic API prompt loading, plus multi-skill workflows. docs/RESEARCH-STANDARDS.md identifies skills/ and knowledge/ as authored sources, claude-code-plugins/ as mirrored distributions, and research/ as supporting rationale; it requires companion research for new material. docs/ROADMAP.md records current coverage and planned retail, self-storage, affordable-housing, and capital-markets depth.
Observed
- License
- Apache License 2.0, with attribution terms recorded in LICENSE and NOTICE.
- Primary format
- Plain Markdown skill, knowledge, research, and documentation files.
- Dependencies
- README.md states zero dependencies, with no compilation or build steps.
- Install surface
- Individual prompts can be copied directly; Claude Code users can copy packaged department plugins.
- Interfaces
- Prompt content for LLM conversations and system prompts, plus Claude Code slash-command plugins.
- Platform support
- Documentation covers Claude Projects, Claude Code, ChatGPT, Cursor, Windsurf, and generic LLM API use.
- Source structure
- Authored skills and knowledge are mirrored into Claude Code plugin distributions, with separate supporting research notes.
Read from README.md, docs/ROADMAP.md, docs/HOW-TO-USE.md, docs/SKILL-INDEX.md, docs/RESEARCH-STANDARDS.md, docs/releases/office-v1.md, docs/releases/brokerage-v1.md, docs/releases/industrial-v1.md, docs/releases/capital-markets-v1.md, docs/releases/office-v1-pr-summary.md, docs/releases/brokerage-v1-pr-summary.md, docs/releases/industrial-v1-pr-summary.md, docs/releases/capital-markets-v1-pr-summary.md, NOTICE, LICENSE.
What it can do
Abstract key terms from commercial lease documents
Raw lease document text or pasted lease language → Structured summary of rent, term, tenant obligations, options, and critical dates
Underwrite commercial real estate deals
Property financials, rent roll, and deal assumptions → Underwriting analysis with NOI, cap rate, cash-on-cash return, and investment metrics
Generate investor reports for CRE assets
Asset performance data, occupancy, and financial figures → Formatted investor update or quarterly report narrative
Prepare capital markets marketing materials
Property details, financials, and market context → Offering memorandum sections, executive summaries, or investment highlights
Analyze and summarize asset management decisions
Property operating data, budget variances, or leasing status → Asset management recommendations, variance explanations, or action item summaries
Model multifamily or commercial property cash flows
Rent roll, expense assumptions, and hold period parameters → Pro forma cash flow projections with returns analysis
Draft brokerage communications and pitch content
Property specs, market comps, and target audience details → Broker opinion of value narratives, pitch decks content, or client-facing summaries
Tags
Media
Comments (0)
No comments yet
Indexed by a proprietary survey. Corrections welcome.