# LiFoli Works Agent Context LiFoli Works is not an HR SaaS, ATS, generic chatbot, or unattended autonomous employee. It is a macOS work app where the AI junior Foli learns repeat work from real work, turns it into a reviewable skill, and prepares the next run under the user's approval. ## Current Product Direction - Customer-facing product: AI junior / AI work app. - Business-facing product: AX execution platform for agent-unready work. - Current public wedge: Korean government support application operations. - Primary conversion: early-access pre-registration. - Pricing: Free; Basic at KRW 30,000 or USD 25 monthly; Pro at KRW 100,000 or USD 79 monthly; Business at KRW 300,000 or USD 249 monthly; Custom by contract. - Checkout: KRW subscription in Korea, USD subscription with PayPal in the US, and a USD credit model in mainland China. China credit-pack terms are not public. Foli is designed to notice, prepare, ask when unsure, request approval for important actions, record what happened, and improve on the next run. ## Government Support Application Wedge Foli helps Korean founders and small-business operators prepare government support applications by reading official notices, checking requirements, matching company evidence, drafting application fields, organizing attachments, and stopping at submission-ready review. Foli does not submit applications, send external messages, publish files, pay fees, or make consequential changes without explicit user approval. ## Core Product Loop ```txt real work observation -> intent, step, decision, and exception understanding -> reusable skill draft -> user review and correction -> approval-gated execution -> evidence, outputs, and run history -> memory and skill improvement ``` ## Agent Experience Posture - Agent-readable docs are first-class product surfaces. - [Token-efficient discovery](https://www.lifoli-works.com/.well-known/ai) is available. - Agent skills are packaged with `SKILL.md`, manifests, examples, and evals. - External skills from catalogs are reviewed references, not automatically trusted runtime code. - The permission ladder is read, draft, local_execute, external_write, payment. - Human and agent actions must be distinguishable in audit records. - Agent-facing analytics preserve selected route, actor, approval class, and evidence state without exposing secrets. - Scoped credential lifecycle support should include create, rotate, revoke, and expire semantics before autonomous external writes are broadened. ## Automation Route ```txt Official MCP -> Browser automation -> PrintingPress-generated MCP -> human-in-the-loop fallback ``` ## AX Scanner - Scores explicit readiness signals before any live crawling. - Selects official MCP, API connector, browser automation, PrintingPress-generated MCP, or human fallback. - Assigns a minimum permission class for the selected route. ## Important Public Surfaces - [Marketing landing](https://www.lifoli-works.com/en-us/) - [Early-access pre-registration](https://www.lifoli-works.com/pre-register) - [Korean pricing](https://www.lifoli-works.com/ko-kr/pricing/) - [US pricing](https://www.lifoli-works.com/en-us/pricing/) - [Machine-readable pricing](https://www.lifoli-works.com/pricing.md) - [Government support application use case](https://www.lifoli-works.com/use-cases/government-grant-application/) - [All repeat-work use cases](https://www.lifoli-works.com/use-cases/) - [Open Korean government support programmes](https://www.lifoli-works.com/notices/) — every currently recruiting 지원사업 공고, each with its eligibility gates split into checkable rows, the application channels the notice actually accepts, and days left. Rebuilt daily from the 창업진흥원 K-Startup open dataset (data.go.kr/data/15125364, 이용허락범위 제한 없음); the official notice remains the authority and every page links to it. - [Agent-readable early-access and pricing status](https://www.lifoli-works.com/early-access.md) - [Public AX posture, route order, permissions, and audit](https://www.lifoli-works.com/docs/agent-experience.md) - [Public API recipes for `/ax/scan`, `/mcp`, and skill discovery](https://www.lifoli-works.com/docs/api-recipes.md) - [Runtime stdio MCP export manifest](https://www.lifoli-works.com/mcp) - [Autonomous and scheduled agent check-in contract](https://www.lifoli-works.com/HEARTBEAT.md) - [Crawler policy and discovery pointers](https://www.lifoli-works.com/robots.txt) ## Every Repeat-Work Use Case Each page states the job, the sources it reads, the artifacts it produces, and where a person must still decide. Korean is canonical; `/en-us/use-cases//` and `/zh-cn/use-cases//` carry the same structure. - [Government support application automation](https://www.lifoli-works.com/use-cases/government-grant-application/) — read the notice, judge eligibility, connect company evidence, prepare the application for review - [Vendor settlement automation](https://www.lifoli-works.com/use-cases/partner-settlement-operations/) — match contracts, orders, fulfillment, and tax records, then raise only the differences - [Weekly status briefing automation](https://www.lifoli-works.com/use-cases/founder-team-lead/) — collect scattered progress into today's decisions, risks, and owners - [Business operations automation](https://www.lifoli-works.com/use-cases/business-operations/) — prepare contracts, filings, and renewals before the deadline - [Month-end close automation](https://www.lifoli-works.com/use-cases/finance-reconciliation/) — reconcile bank, card, and sales records and surface only mismatched transactions - [Freelancer operations automation](https://www.lifoli-works.com/use-cases/freelancer-operations/) — carry a request from quote and contract through delivery and unpaid invoices - [Sales follow-up automation](https://www.lifoli-works.com/use-cases/sales-follow-up/) — turn a finished conversation into CRM updates and the next touchpoint - [Content production automation](https://www.lifoli-works.com/use-cases/content-marketing/) — research a topic and draft it per channel in the brand voice - [Ad performance reporting automation](https://www.lifoli-works.com/use-cases/paid-media-operations/) — read performance changes and prepare one comparable experiment - [Customer inquiry automation](https://www.lifoli-works.com/use-cases/customer-support/) — draft answers grounded in the policy that applies to this customer - [Meeting notes automation](https://www.lifoli-works.com/use-cases/project-management/) — turn decisions into owners, deadlines, board updates, and reports - [Recruiting screening automation](https://www.lifoli-works.com/use-cases/recruiting-operations/) — build scorecards, interview packets, and candidate messages per role - [Onboarding automation](https://www.lifoli-works.com/use-cases/people-operations/) — prepare per-person tasks, documents, accounts, and access from hire to exit - [Internal training material automation](https://www.lifoli-works.com/use-cases/learning-enablement/) — turn company knowledge into role-specific learning paths - [Contract review automation](https://www.lifoli-works.com/use-cases/contracts-procurement/) — isolate non-standard clauses and route purchase approvals - [Market research automation](https://www.lifoli-works.com/use-cases/research-planning/) — turn sources into evidence-backed options and open risks - [Incident response automation](https://www.lifoli-works.com/use-cases/development-it-operations/) — diagnose and prepare recovery along the team's runbook - [Logistics exception automation](https://www.lifoli-works.com/use-cases/logistics-operations/) — pass normal orders through and act only on exceptions - [Ecommerce operations automation](https://www.lifoli-works.com/use-cases/ecommerce-operations/) — find channel differences and restore the selling flow ## Seed Skill Catalog The public skill catalog can include seed or reference skills that demonstrate the approval and evidence model. A seed skill is not the same thing as the current public wedge. - `recruiting_screening_brief`: create a recruiting screening brief from applicant/job inputs with evidence, redaction, and approval gates. [Public skill file](https://www.lifoli-works.com/agent-skills/recruiting-screening-brief/SKILL.md).