# Personal AI Operating Site — full public context

> Curated public bundle for AI agents. This is NOT a full site dump.
> Maintained by Wonbeen Lee (wbeen). Only visibility: public | generalized content is included.

## 1. Project definition

A personal knowledge product for AI-era product work: essays, copy-ready
prompts and templates, lab notes, projects, skills, and references — published for
both human readers and AI agents from one typed content source.

## 2. Public positioning

- Not a portfolio, not a blog, not an AI SaaS landing page.
- Reusable workflows over one-off prompts.
- A public layer of a private knowledge system.

## 3. Content type definitions

- articles: Long-form essays and operating notes.
- prompts: Copy-ready prompts with declared inputs, outputs, and safety notes.
- templates: Document and workflow templates with section lists.
- labs: Experiment records: hypothesis, setup, result, next iteration.
- projects: Running experiments with public artifacts and related notes.
- references: External sources with short commentary on why they matter.
- notes: Short atomic evergreen notes.
- skills: Reusable AI operating capabilities with triggers, inputs, outputs, guardrails, and verification criteria.
- workflows: Editorial routing objects that connect a current work state to an ordered next-action path.

## 4. Route map

- / (home)
- /articles and /articles/[slug]
- /prompts and /prompts/[slug] (not yet live — v0.4+)
- /templates and /templates/[slug] (not yet live — v0.4+)
- /labs and /labs/[slug]
- /projects and /projects/[slug]
- /references and /references/[slug] (not yet live — v0.4+)
- /notes and /notes/[slug]
- /skills and /skills/[slug]
- /workflows (canonical Router; records are exposed through /workflows.json, /workflows.md, and the Markdown mirror)
- /agents and /agents/[id] (public Agent Society showcase and canonical identity detail pages)
- /knowledge-map and /content-graph.json
- /llms.txt, /llms-full.txt, /sitemap.xml, /rss.xml, /md/[type]/[slug].md

Note: /articles, /notes, /projects, /labs, /skills, the /workflows Router, and /agents are live.
Collections without HTML detail routes remain available through the Markdown mirror layer.

## 5. Public/private boundary

- Every item is classified: public | generalized | review | private-only.
- Only public and generalized content ships to any route or export.
- private-source and private-only content never appears in this bundle.

## 6. Knowledge graph

- /knowledge-map explains the public relationship layer for humans.
- /content-graph.json exposes stable nodes and edges for agents.
- Nodes include id, title, type, url, summary, description, topics, and tags.
- Edges include source, target, and relation.

## 7. Public content summaries

### articles

- A Prompt Library Is Not Enough — Claims that prompt dumps decay; durable value comes from workflows with typed inputs/outputs, use cases, and safety notes.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/a-prompt-library-is-not-enough
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/a-prompt-library-is-not-enough.md
- My journey of bringing AI into Product Planning (Part 2) — Hub, pipeline, and governance — Introduces a three-layer operating model for AI-assisted product planning — Hub for multi-tool reality, Pipeline for standardized deliverables, Governance for repeatable team gains.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/ai-product-planning-hub-pipeline-governance
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/ai-product-planning-hub-pipeline-governance.md
- My journey of bringing AI into Product Planning (Part 5) — What this changed about the PM role — Closes the series by arguing that AI relocates product judgment rather than removing it, shifting the PM role toward context design, workflow design, and lightweight governance.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/ai-product-planning-pm-role
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/ai-product-planning-pm-role.md
- My journey of bringing AI into Product Planning (Part 1) — Why AI adoption forced a process redesign — Argues that AI makes drafting cheap and misalignment expensive, and that adopting AI in product planning requires redesigning the process (deliverables + quality gates), not just picking a tool.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/ai-product-planning-process-redesign
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/ai-product-planning-process-redesign.md
- My journey of bringing AI into Product Planning (Part 3) — Repeatable use cases inside the workflow — Describes four AI use cases in product planning (documentation, research support, prototyping, review) that mattered because they connected to the surrounding workflow rather than existing as one-off outputs.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/ai-product-planning-repeatable-use-cases
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/ai-product-planning-repeatable-use-cases.md
- My journey of bringing AI into Product Planning (Part 4) — From personal prompts to team capability — Argues that scaling AI in product planning requires moving from individual prompt skill to reusable team capability, built on clear inputs, stable outputs, shared standards, and interconnected workflow components.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/ai-product-planning-team-capability
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/ai-product-planning-team-capability.md
- The rollout starts after setup — Published LinkedIn post on Claude adoption in product planning, emphasizing that AI systems need source hygiene, skill improvement, output review, team sharing, and lightweight governance after setup.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/ai-rollout-after-setup
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/ai-rollout-after-setup.md
- Before Asking AI to Change Anything: An Evidence-first Inspect, Verify, and Audit Workflow — Guide to an evidence-first AI work sequence: inspect the current state, turn requirements into observable checks, make the smallest justified change, verify with tool evidence, and stop or escalate when evidence is unavailable. Links to Prompt Hotbar prompts and the coding-agent workflow.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/evidence-first-ai-workflow
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/evidence-first-ai-workflow.md
- Fable 5: Audit the Checklist That Approved the Work — Published LinkedIn post on using Fable 5 to audit completed work, distinguish recorded and reflected evidence from execution verification, and challenge the checklist itself.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/fable-5-audit-the-checklist
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/fable-5-audit-the-checklist.md
- From Prompt Reuse to Personal Retrieval: Preference Memory Without Accounts — Prompt Hotbar v0.6.0 case-note: browser-local positive preferences (Like / Love this) and My Picks provide personal retrieval without accounts or server-side profile state. Personal prompt content and local prompt IDs stay on-device; privacy-bounded PostHog events measure curated preference changes, My Picks use, and downstream product actions. A backend remains intentionally out of scope until personalization must cross browser or device boundaries.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/prompt-hotbar-local-first-preference-layer
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/prompt-hotbar-local-first-preference-layer.md
- From Prompt Clipboard to Local MCP: Testing Prompt Hotbar Inside Claude Code — Proof-of-work note on a private Prompt Hotbar MCP pilot in Claude Code. A local stdio server exposed five read-only tools for prompt search, prompt retrieval, recommendation, workflow search, and workflow retrieval. The pilot returned reviewed public-safe catalog data only: 23 prompts and 3 workflows. npm publication remained intentionally withheld.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/prompt-hotbar-local-mcp-pilot
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/prompt-hotbar-local-mcp-pilot.md
- Why I'm Building a Personal AI Operating Site — Argues that personal sites should evolve from portfolios into typed, AI-readable knowledge products with explicit public/private boundaries.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/why-im-building-a-personal-ai-operating-site
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/why-im-building-a-personal-ai-operating-site.md
- Why YouTube Opened PiP but Kept Background Play Premium — Analyzes YouTube's decision to free Picture-in-Picture while keeping background play Premium, arguing the split reflects a monetization boundary between a visual attention surface and an audio-first paid habit.
  canonical: https://personal-ai-ops.vercel.app/wbeen/articles/why-youtube-opened-pip-kept-background-play
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/articles/why-youtube-opened-pip-kept-background-play.md

### prompts

- Agentic Handoff Brief — Prompt that produces a structured handoff brief (goal, state, constraints, next actions) for continuing work across agents or sessions.
  canonical: https://personal-ai-ops.vercel.app/wbeen/prompts/agentic-handoff-brief (not yet live — v0.4+)
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/prompts/agentic-handoff-brief.md
- Conversation-to-Source Extractor — Prompt that extracts reusable assets from AI chat exports as classified markdown sources with public/review/private sensitivity labels.
  canonical: https://personal-ai-ops.vercel.app/wbeen/prompts/conversation-to-source-extractor (not yet live — v0.4+)
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/prompts/conversation-to-source-extractor.md

### templates

- AI-Readable Site IA Planner — Planning template for designing dual-layer site IA: human routes plus AI-readable entry points such as llms.txt and markdown mirrors.
  canonical: https://personal-ai-ops.vercel.app/wbeen/templates/ai-readable-site-ia-planner (not yet live — v0.4+)
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/templates/ai-readable-site-ia-planner.md
- Public/Private Boundary Checker — Pre-publish checklist template enforcing the public/private boundary: no company internals, client names, or personal data in public content.
  canonical: https://personal-ai-ops.vercel.app/wbeen/templates/public-private-boundary-checker (not yet live — v0.4+)
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/templates/public-private-boundary-checker.md

### labs

- Chat Conversation Mining for Public Knowledge — Lab note on mining AI chat exports for publishable assets using an extraction prompt with built-in sensitivity classification.
  canonical: https://personal-ai-ops.vercel.app/wbeen/labs/chat-conversation-mining-for-public-knowledge
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/labs/chat-conversation-mining-for-public-knowledge.md
- From Canonical Profile to Essence: Designing a Motion-Ready Identity System for AI Agents — Lab note on building motion-ready Essence identities for fourteen AI agents. Covers shape-first design, autonomous gaze/blink, 2.5D depth, cursor-local interaction, mobile/reduced-motion behavior, and refinement failures such as Mina cropping, Hedwig pseudo ears, Iris prism occlusion, Vesta silhouette drift, and detail/index motion drift.
  canonical: https://personal-ai-ops.vercel.app/wbeen/labs/from-canonical-profile-to-essence
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/labs/from-canonical-profile-to-essence.md
- Multiplatform Context Handoff — Lab note on testing a workflow and protocol for preserving AI coding session context across ChatGPT, Claude Code, Codex CLI, and Antigravity.
  canonical: https://personal-ai-ops.vercel.app/wbeen/labs/multiplatform-context-handoff
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/labs/multiplatform-context-handoff.md
- Personal AI Operating Site 1-day Prototype — Lab note testing whether a personal site reads as a knowledge product; prototype evaluated externally, then rebuilt as an Astro scaffold.
  canonical: https://personal-ai-ops.vercel.app/wbeen/labs/personal-ai-operating-site-one-day-prototype
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/labs/personal-ai-operating-site-one-day-prototype.md
- Public Knowledge Router: Turning Raw AI Conversations into Publishable Knowledge Assets — Interactive lab prototype demonstrating the public/private routing pipeline: raw input to sanitized summary, frontmatter, canonical page, markdown mirror, and llms-full.txt entry. Static, no backend.
  canonical: https://personal-ai-ops.vercel.app/wbeen/labs/public-knowledge-router
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/labs/public-knowledge-router.md

### projects

- Personal Agent Society — A bounded multi-agent operating model — Public architecture and product-thinking record for a bounded multi-agent operating model. Covers role division, Personal/Work boundaries, Canonical Profile → Essence → runtime lineage, relationship-versus-authority separation, and public/private limits.
  canonical: https://personal-ai-ops.vercel.app/wbeen/projects/personal-agent-society
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/projects/personal-agent-society.md
- Personal AI Operating Site — Personal knowledge product built from one public-safe typed content source with separate human and AI reading paths, deterministic Markdown/llms/OKF derivatives, a richer typed relationship graph, Career progressive disclosure, and an enforced public/private publishing boundary.
  canonical: https://personal-ai-ops.vercel.app/wbeen/projects/personal-ai-operating-site
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/projects/personal-ai-operating-site.md
- Prompt Hotbar — Keyboard-first retrieval workspace with nine configurable slots, 72 built-in public-safe prompts, 11 preset and Workflow Sets, browser-local personal prompts and positive preferences, reviewed Prompt/Workflow/LAB Discovery pages, six LAB challenges, and a 134-card Toolbox spanning Claude Code, Codex, Antigravity, Hermes Agent, and OpenCode. Toolbox references pair copyable commands or configuration recipes with official provenance, review dates, and bounded risk cues; low-frequency Install / Setup lives behind Provider utilities rather than permanent cards, and unsupported symmetric commands are omitted instead of invented. Prompt Hotbar copies instructions and references but does not execute prompts, commands, installers, or configuration changes. A private local read-only MCP pilot separately tested reviewed in-client retrieval without accessing browser-local personal state.
  canonical: https://personal-ai-ops.vercel.app/wbeen/projects/prompt-hotbar
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/projects/prompt-hotbar.md
- Universal Context Handoff — Local-first CLI and assistant skill for preserving AI coding session context across repositories and environments (Claude Code, Codex, Antigravity legacy JSONL).
  canonical: https://personal-ai-ops.vercel.app/wbeen/projects/universal-context-handoff
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/projects/universal-context-handoff.md

### references

- The llms.txt Pattern — Reference on the llms.txt convention: a curated markdown entry point for LLMs, which this site adopts as its agent-facing index.
  canonical: https://personal-ai-ops.vercel.app/wbeen/references/llms-txt-pattern (not yet live — v0.4+)
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/references/llms-txt-pattern.md

### notes

- Mapping confidence levels in public knowledge bases — Observation note: Explicit confidence mapping (low/medium/high) is a crucial signal for both humans and agents in public knowledge bases.
  canonical: https://personal-ai-ops.vercel.app/wbeen/notes/confidence-levels-in-public-kb
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/notes/confidence-levels-in-public-kb.md
- Designing agents with boundaries, not just personas — Principle note: multi-agent design should model role, environment, authority, context, relationship, and handoff separately. Uses Mina/Miro and Hedwig/Haechi as public examples of related identity without shared authority or runtime permission.
  canonical: https://personal-ai-ops.vercel.app/wbeen/notes/designing-agents-with-boundaries-not-just-personas
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/notes/designing-agents-with-boundaries-not-just-personas.md
- Minimum structure for useful AI context — Concept note: AI context alignment requires minimal reusable structures (inputs, outputs, states) rather than raw text dumps.
  canonical: https://personal-ai-ops.vercel.app/wbeen/notes/minimum-structure-for-useful-ai-context
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/notes/minimum-structure-for-useful-ai-context.md
- Reusable workflows over one-off prompts — Principle note: durable AI leverage comes from workflows with declared inputs/outputs, not from collections of prompt text.
  canonical: https://personal-ai-ops.vercel.app/wbeen/notes/reusable-workflows-over-prompts
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/notes/reusable-workflows-over-prompts.md

### skills

- Agent Handoff — A public-safe operating skill for moving work between AI agents or sessions while preserving decisions, evidence, constraints, and the next executable action.
  canonical: https://personal-ai-ops.vercel.app/wbeen/skills/agent-handoff
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/skills/agent-handoff.md
- Conversation-to-Source — A public-safe operating skill for extracting reusable source candidates from AI conversations and routing ambiguous material to human review.
  canonical: https://personal-ai-ops.vercel.app/wbeen/skills/conversation-to-source
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/skills/conversation-to-source.md
- Evidence-first Review — A public-safe operating skill for reviewing AI-assisted work with observable checks, evidence, and explicit unresolved risks.
  canonical: https://personal-ai-ops.vercel.app/wbeen/skills/evidence-first-review
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/skills/evidence-first-review.md
- Humanize Loop — A public-safe UI humanization operating skill that classifies context, routes to at most one primary Humanizer plus one validator, preserves product intent, and accepts no intervention when the UI is already intentional.
  canonical: https://personal-ai-ops.vercel.app/wbeen/skills/humanize-loop
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/skills/humanize-loop.md
- Public-safe Publishing — A public-safe publishing skill for verifying boundary, metadata, links, generated mirrors, and live evidence before release.
  canonical: https://personal-ai-ops.vercel.app/wbeen/skills/public-safe-publishing
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/skills/public-safe-publishing.md
- Scope & Acceptance Criteria — A public-safe planning skill for turning a goal into bounded scope and observable acceptance criteria before work begins.
  canonical: https://personal-ai-ops.vercel.app/wbeen/skills/scope-acceptance-criteria
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/skills/scope-acceptance-criteria.md
- Session Handoff — A vendor-neutral continuity skill for near-limit AI work. It uses PREPARE and EVACUATE stages, recoverable source checkpoints, a sanitized handoff capsule, one-owner execution leases, verified takeover, and an optional emergency-seat return cycle.
  canonical: https://personal-ai-ops.vercel.app/wbeen/skills/session-handoff
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/skills/session-handoff.md

### workflows

- Delegate a bounded AI task — Workflow for delegating a bounded AI task: define scope and acceptance criteria, retrieve an appropriate prompt or invocation surface, and perform evidence-first review.
  canonical: https://personal-ai-ops.vercel.app/wbeen/workflows/#workflow-delegate-bounded-ai-task
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/workflows/delegate-bounded-ai-task.md
- Resume or transfer ongoing work — Workflow for resuming or transferring ongoing work: create a bounded handoff, continue from its next action, and verify the resumed result.
  canonical: https://personal-ai-ops.vercel.app/wbeen/workflows/#workflow-resume-or-transfer-ongoing-work
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/workflows/resume-or-transfer-ongoing-work.md
- Turn a working conversation into a public asset — Workflow for turning a working AI conversation into a public asset: extract reusable material, review the public/private boundary, and publish a generalized result.
  canonical: https://personal-ai-ops.vercel.app/wbeen/workflows/#workflow-turn-conversation-into-public-asset
  mirror: https://personal-ai-ops.vercel.app/wbeen/md/workflows/turn-conversation-into-public-asset.md

## 8. Tag glossary

`AI Product Ops` · `UI design` · `acceptance` · `acceptance-criteria` · `agent-society` · `agent-ux` · `agentic-ai` · `agentic-workflows` · `ai-adoption` · `ai-coding` · `ai-fluency` · `ai-in-product` · `ai-product-planning` · `ai-readability` · `ai-readable` · `ai-workflow` · `alignment` · `analytics` · `anti-ai-slop` · `antigravity` · `astro` · `audit` · `authority` · `boundaries` · `boundary` · `boundary-safe` · `case-study` · `checklist` · `checkpoint` · `clarity` · `classification` · `claude` · `claude-code` · `cli` · `codex-cli` · `coding-agents` · `confidence` · `context` · `context-handoff` · `continuation` · `continuity` · `conversation` · `delegation` · `developer-tools` · `discovery` · `essence` · `evidence` · `evidence-first` · `experiment` · `extraction` · `fable-5` · `frontend` · `governance` · `handoff` · `humanize` · `identity-system` · `inspection` · `interaction-design` · `interactive` · `iteration` · `knowledge-architecture` · `knowledge-capture` · `knowledge-ops` · `knowledge-product` · `library` · `llms-txt` · `local-first` · `machine-readable` · `mcp` · `metadata` · `monetization` · `motion-design` · `multi-agent` · `my-picks` · `okf` · `open-source` · `operating-design` · `operating-model` · `operating-structure` · `personal-kb` · `personalization` · `planning` · `planning-workflow` · `pm-role` · `preference-memory` · `principle` · `privacy` · `process-redesign` · `product-design` · `product-evolution` · `product-management` · `product-ops` · `product-planning` · `product-strategy` · `productivity` · `prompt-hotbar` · `prompt-reuse` · `prototype` · `public-safe` · `publishing` · `publishing-boundary` · `publishing-pipeline` · `quality` · `quality-gate` · `react` · `read-only` · `release` · `return-handoff` · `reusability` · `reuse` · `review` · `scope` · `scope-control` · `session-continuity` · `site-structure` · `source` · `standards` · `stdio` · `structure` · `system-design` · `takeover` · `team-capability` · `thesis` · `trust` · `typescript` · `use-cases` · `user-behavior` · `verification` · `visual-identity` · `vite` · `workflow` · `workflow-design` · `workflows`