بنقرة واحدة
act-global-infrastructure
يحتوي act-global-infrastructure على 22 من skills المجمعة من Acurioustractor، مع تغطية مهنية على مستوى المستودع وصفحات skill داخل الموقع.
Skills في هذا المستودع
Read ACT's energy-orbit relational CRM — where a person sits (supporter ring vs community constellation), their evidence or owes, flags (violation/dupe/ghost/uncaptured), and the gated next move. Read-only by default; never auto-writes. Use when user says "orbit", "/orbit", "where does <person> sit", "show me the orbit", "is <person> on the wrong lane / a community-line violation", "who's uncaptured", or wants to read or act on the relationship system before any GHL write.
BAS quarter preparation, receipt acquittal, and learning loop for ACT. Use for BAS prep, missing receipt hunts, vendor reconciliation, and quarterly retrospectives. Accumulates patterns and learnings across quarters so each BAS cycle gets cleaner.
Drives the NAB Visa
Query actual database schema before writing any query, migration, or API endpoint. Prevents wrong column names — the
Verify Supabase connection, TypeScript health, and schema before any database or implementation work. Use when starting a session, before migrations, or when something feels off.
Review and tag untagged Xero transactions / bank lines / invoices with project codes — the cross-project spend & income alignment workflow. Use when user says "tag transactions", "fix transaction tagging", "improve transaction coverage", or "align spend/income across projects".
OCAP consent gate. Run BEFORE publishing, syndicating, or externally sharing ANY storyteller content — a story, quote, name, photo, or video — to wiki story pages, newsletters, Empathy Ledger syndication, social, funder decks, or the website. Grills the consent provenance, verifies names and place names, and BLOCKS publish if the audit trail is incomplete or anything is unverified. Use whenever you are about to make community or storyteller content visible outside ACT.
Capture a contemporaneous R&D record at the close of engineering work that resolved genuine technical uncertainty — novel architecture, an experimental integration, or an approach whose outcome was not knowable in advance. Appends to the FY26 R&D activity register for the 43.5% R&D Tax Incentive. Do NOT use for routine work — CRUD, config, copy, dependency bumps, or bug fixes with a known solution are not R&D.
ACT brand alignment for all ecosystem projects. Use for ANY work on ACT sites, projects, content, design, or strategy. Understands ACT identity, LCAA method, dual-entity structure, all projects, voice/tone, and visual language.
ACT Wikipedia — ingest sources, query knowledge, lint for gaps, and manage the LLM-compiled wiki at wiki/. Use when user says "wiki", "ingest", "add to wiki", "wiki query", "wiki lint", "knowledge base", or wants to add/query/maintain the ACT knowledge base.
Skip planning, query schema, implement directly. Use when you want code not plans. Say "/implement [feature description]" to go straight to building.
Run the GHL alignment script to review and tag opportunities with project codes. Use when user says "align ghl", "tag opportunities", or "link opportunities to projects".
Check ecosystem data quality and alignment scores across all data sources. Use when user says "ecosystem health", "data quality", "alignment scores", or "how aligned is the data".
Research advisor for ACT business setup — Pty Ltd registration, accountants, insurance, R&D tax incentive, family trusts, compliance, and migration. Use when the user asks about any business setup topic, costs, providers, or next steps.
Deploy to Vercel with pre-flight checks, env var parity, build verification, and post-deploy health checks. Use when user says "deploy", "push to production", "ship it", or after completing a feature.
Start local development environment — dev servers, tunnels for webhook testing, health checks, and log tailing. Use when user says "start dev", "local dev", "spin up", "test locally", "tunnel", or wants to test webhooks.
**Obsidian Vault Manager**: Create, edit, and organise Obsidian-compatible markdown notes with proper wiki-links, YAML frontmatter, tags, callouts, dataview queries, and vault structure. Use this skill whenever the user mentions Obsidian, vault, wiki-links, [[backlinks]], note-taking, second brain, PKM, knowledge management, zettelkasten, or wants to create markdown files that work in Obsidian. Also trigger when creating documentation, project notes, meeting notes, daily notes, or MOCs (Maps of Content) that should be Obsidian-compatible. This skill ensures all markdown output uses Obsidian-native syntax rather than generic markdown. Even if the user doesn't explicitly mention Obsidian, trigger when the context suggests notes should be vault-compatible (e.g. creating linked knowledge, structured notes with metadata, or files going into a docs/ folder alongside other .md files).
Sprint planning, daily standups, health monitoring, and issue automation for ACT ecosystem development.
Autonomous content creation, review, and multi-platform publishing for ACT ecosystem. Includes Empathy Ledger Content Hub for photography, videography, and storytelling with AI-powered tagging and ecosystem syndication.
Secure management of environment variables and secrets across ACT projects. Use for secret audits, rotation, setup, sync, and security scans.
Strategic advisor for GoHighLevel CRM implementation across ACT projects. Use for pipeline design, workflow automation, integrations, and CRM strategy.
Ralph - Long-Running Task Agent. Autonomous AI that works through PRD tasks iteratively, with access to full ACT ecosystem knowledge across all 7 codebases.