菜单
County-by-county scrape of construction builders (GCs, home builders, remodelers). Dedups across counties, classifies vs trades/suppliers, enriches via Firecrawl websites + Facebook pages. Outputs one master CSV per state.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Generate a polished post-workshop follow-up package for a Ruh AI client engagement: an HTML takeaway document, a PDF export, and a short cover email in Luke's voice. Use this skill whenever the user has just finished a Ruh AI workshop with a client (typically a Claude Code / AI enablement session led by Jesse and/or Luke) and needs to send the attendees a follow-up. Input is a transcript (file path or pasted text) and a recording link; the skill auto-discovers attendees from Google Calendar when available, does light company research from recipient email domains, runs a six-pass extraction over the transcript to capture Jesse-isms, attendee Q&A, identified use cases, prompts shared in chat, commitments, and personalization hooks — then renders the artifacts into the client's working directory under `Workshop-N/`. The output matches the house style established for Taurus Builders and Dondlinger. Triggers on: "workshop follow-up", "workshop followup", "workshop takeaway", "workshop recap", "post-workshop email"
Generate professional consulting engagement letters as Word documents (.docx). Use this skill whenever the user asks for an engagement letter, consulting agreement, statement of work for consulting or coaching, training engagement proposal, or any letter that outlines the terms of a consulting/coaching/training engagement with a client. Also trigger when the user mentions 'SOW', 'letter of engagement', 'consulting proposal', 'coaching agreement', or wants to formalize a client conversation into a professional document for signature. Produces a polished, branded Ruh AI engagement letter with outcome-based framing, pricing, guarantee, and signature blocks. Triggers on: "engagement letter", "consulting agreement", "SOW", "statement of work", "letter of engagement", "consulting proposal", "coaching agreement", "put together a letter", "send them something to sign", "formalize this engagement".
Turn a deal folder (transcripts, proposals, MSA, POC outputs, skill files) into a structured developer onboarding document that brings a Ruh.AI engineering team from zero context to productive in ~4 hours. The output covers: who the client is, industry primer, current-state workflow, agent product spec, technical architecture, training pipeline, commercial + SLA envelope, risks and gotchas, and prioritized open questions. Use this skill whenever a deal has closed (verbal or signed) and the dev team needs to assume the build. Delivers markdown, Word (.docx), and PDF outputs in the same folder. Triggers on: "dev onboarding doc", "dev team onboarding", "engineering handoff", "bring dev team up to speed", "developer onboarding", "handoff doc", "turn this folder into a dev doc", "build context doc", "engineering brief", "onboard the engineers", "scaffold for the build team", "dev team context document".
Enrich the memory-kb contact registry into compounding person pages. For each contact with new activity, build or update knowledge/concepts/people/{slug}.md by scanning daily logs, inferring company from email domain, and optionally WebSearching for role/company details. Idempotent, batched, wikilinks into existing concepts. Triggers on: "enrich contacts", "update person pages", "build contact pages", "refresh contact registry", "contact enrich", "/memory-contact-enrich".
Generate enterprise Master Services Agreements (MSAs) with accompanying proposals as Word documents (.docx). Use this skill whenever the user asks to create an MSA, master services agreement, service agreement, SaaS agreement, platform license agreement, or AI agent deployment contract. Also trigger when the user wants to formalize a client engagement that involves ongoing platform access, agent deployment, or technology licensing. This skill ingests meeting transcripts, presentations (.pptx), notes, and directory context to produce a complete MSA with scope of work, pricing, IP protections, performance milestones, and data portability provisions. Produces a polished, branded Ruh AI MSA ready for client signature. Triggers on: "MSA", "master services agreement", "service agreement", "create an MSA", "put together a contract", "draft the agreement", "formalize this deal", "SaaS agreement", "platform license", "agent deployment contract", "write up the MSA", "generate the contract".
Sync shared skills from the SharedSkillsLuke GitHub repo. Checks the repo for changes, pulls the latest, and copies any new or updated skill folders into ~/.claude/skills/. Also pushes any local skill changes (for skills tracked in the repo) back to GitHub. Designed to run daily as a scheduled task to keep Jesse and Luke in sync. Triggers on: "sync skills", "sync shared skills", "pull shared skills", "push shared skills", "update shared skills", "skill sync".
| name | construction-sweep |
| description | County-by-county scrape of construction builders (GCs, home builders, remodelers). Dedups across counties, classifies vs trades/suppliers, enriches via Firecrawl websites + Facebook pages. Outputs one master CSV per state. |
| triggers | ["/construction-sweep","sweep construction","scrape builders"] |
Purpose: Build a state-level database of construction builders by running a repeatable per-county pipeline: Apify Google Maps scrape → cross-county dedup → rules classifier → Firecrawl + FB enrichment → append to master CSV.
First time on this machine? Read SETUP.md in this skill dir FIRST. Verify prereqs before invoking.
User types one of:
/construction-sweep <State> — pick next undone county, run full pipeline/construction-sweep <State> <County> — run specific county/construction-sweep <State> --status — show progress, stats, next county/construction-sweep <State> --bootstrap — initialize state folder (idempotent)/construction-sweep <State> --dry-run — confirm next county + est cost, no API callsState folder defaults to ./<State>/ in cwd. Override with --dir <path>.
Each step writes a checkpoint file so a crash doesn't restart from scratch. Resume by re-running same county — skill detects completed stages and skips.
| # | Stage | Script | Input | Output | Wait |
|---|---|---|---|---|---|
| 1 | Bootstrap state | scripts/bootstrap_state.py <state> <dir> | state name | master.csv + seen.json + processed.txt + counties.json | — |
| 2 | Pick next county | scripts/pick_next_county.py <dir> | state dir | county name | — |
| 3 | Confirm w/ user | (inline) | — | yes/no | user |
| 4 | Scrape GMaps | Apify MCP compass/crawler-google-places async | county + state | runId | 5-15 min poll |
| 5 | Pull dataset | Apify API /v2/datasets/<id>/items | datasetId | <county>_raw.json | — |
| 6 | Dedup county | scripts/dedup_county.py <county> <raw> <dir> | raw.json | <county>_new.json + updated seen.json + master backfills | — |
| 7 | Classify | scripts/classify_county.py <county> <dir> | new.json | <county>_classified.json | — |
| 8 | Prep enrichment | scripts/prep_enrichment.py <county> <dir> | classified.json | <county>_fc_urls.json + <county>_fb_input.json | — |
| 9a | FC batch submit | scripts/submit_firecrawl.py <county> <dir> | fc_urls.json | FC batchId | 3-8 min poll |
| 9b | FB batch submit | Apify MCP apify/facebook-pages-scraper async | fb_input.json | Apify runId | 3-8 min poll |
| 10a | FC pull | scripts/pull_firecrawl.py <county> <batchId> <dir> | batchId | <county>_fc.json | — |
| 10b | FB pull | Apify API dataset pull | runId | <county>_fb.json | — |
| 11 | Append | scripts/append_county.py <county> <dir> | all above | master.csv row per KEEP + seen.json update | — |
| 12 | Log | append to processed.txt | — | — | — |
~/.firecrawl_key (chmod 600). Get from firecrawl.dev.~/.claude.json. Skill uses mcp__Apify__call-actor + public /v2/actor-runs/<id> endpoint (no token needed for status polls).If either missing, abort + tell user what to install. SETUP.md has exact commands.
Use ScheduleWakeup tool for polling:
/v2/actor-runs/<runId> for SUCCEEDEDDo NOT poll in a tight loop. Each poll burns context. Sleep ≥60s between checks.
| Symptom | Fix |
|---|---|
| Apify scrape fails or 0 items | Check input, retry once. If still fails, flag county + skip. |
| Firecrawl "we do not support this site" | URL in blocklist. Check data/blocklists.json, strip bad URL, resubmit. |
| FC pagination returns <expected | Bump max_pages to 100 in pull_firecrawl.py. |
zsh: read-only variable: status | Rename var to st. |
| APIFY_TOKEN empty | Use unauth'd /v2/actor-runs/<id> for status. Only the MCP call-actor needs auth, MCP handles it. |
| NoneType on .lower() | Add or '' guards. |
| Classifier rejects obvious builder | Check if name has "Builders"/"Custom Homes"/"Construction" — manual rescue and add rule to classifier. |
Full troubleshooting in SETUP.md.
Per county (rural): ~$0.50-0.80 Per county (metro): ~$1.00-1.50 Per state (40-50 counties): ~$25-50
Breakdown:
27 columns, one row per unique builder. Key columns:
business_name, counties (pipe-sep, e.g. "Kootenai|Bonner")phone, website, email_primary, all_emails, facebook, instagram, linkedingoogle_rating, google_reviews_countyears_in_business, license_numbers (pipe-sep)services_summary, scope (residential/commercial/civil/mixed)category_primary, categories_all, classification_reasonplace_id (dedup key)Full schema: SETUP.md §"Master CSV schema".
data/counties_by_state.json--bootstrap <State> which prompts for county list OR pulls from census FIPS (future)Master CSV is self-contained. Drop <State>/master.csv into Sheets or CRM. counties field pipe-separated for easy splitting.
For teammate handoff: zip <State>/ folder — includes master + seen index + processed list, so another teammate can resume the sweep from where it stopped.
When skill improves (classifier rules, blocklists, scripts):
cd ~/.claude/skills/construction-sweep && git pullCHANGELOG.md for breaking changesSee SETUP.md for install/update workflows.