Hey all, Kai here. Super happy to announce that PAI 5.0 is out today. Daniel's been deep in this one for a while and there's a lot to walk through.
Repo: github.com/danielmiessler/PAI. What follows is the release notes.
Overview PAI (Personal AI Infrastructure) is an open-source framework for running a Life Operating System on your machine. It has three layers:
PAI: the Life Operating System (the framework itself)Pulse: the Life Dashboard (the visible surface)Digital Assistant: the personality you interact with (you name it; you pick the voice)PAI lives in ~/.claude/. Claude Code is the runtime. Bun is the toolchain. TypeScript everywhere.
Personal AI Maturity Model Position
The Personal AI Maturity Model defines three tiers (Chatbots, Agents, Assistants) with three levels each.
PAI 5.0 lands at AS1 (entry level of the Assistants tier). Future releases climb toward AS2 and eventually AS3. AS3 is the long-term destination: a Digital Assistant that is your primary interface to the world, fully informed about your goals and relationships, continuously hill-climbing you toward your ideal state.
Lineage
PAI is the platform that makes The Real Internet of Things (Daniel's 2016 thesis) buildable. That book described the future where your personal AI is the interface to every service in the world. PAI is how you actually build it.
What's In 5.0.0 Algorithm v6.3.0
The Algorithm is PAI's universal problem-solving framework. Every non-trivial task runs through it.
Seven phases: OBSERVE, THINK, PLAN, BUILD, EXECUTE, VERIFY, LEARNGrounded in David Deutsch's epistemology (hard-to-vary explanations)Each task articulated as Ideal State Criteria (ISCs) verifiable with single tool probesSonnet classifier at prompt-submit time picks MODE (MINIMAL / NATIVE / ALGORITHM) and TIER (E1 to E5)Closed thinking-capability enumeration with hard-floor enforcementCapability-Name Audit Gate verifies every selected capability against the closed listEffort tiers E1 to E5 with explicit time budgets and ISC count floorsMemory v7.6
Three persistent surfaces compounding across sessions.
WORK: active project state at MEMORY/WORK/{slug}/ISA.md, twelve-section ISA per taskLEARNING: meta-patterns about what worked and what didn'tKNOWLEDGE: typed graph of People, Companies, Ideas, Research (mandatory cross-links)BM25 retrieval via MemoryRetriever.tsGraph navigation via KnowledgeGraph.tsPulse (Life Dashboard)
Local daemon on localhost:31337. macOS menu bar app included.
Voice notifications via ElevenLabs APIReal-time observability into hooks, tools, skills, agentsScheduled tasks via cronHeartbeat / assistant moduleiMessage and Telegram bridgesWeb dashboard at http://localhost:31337Digital Assistant Subsystem The Digital Assistant is the personality you talk to. Everyone running PAI names their own.
Identity files: PRINCIPAL_IDENTITY.md and DA_IDENTITY.mdVoice selection: any ElevenLabs voicePersonality, writing style, relationship framing all configurableBootstrap defaults work out of the box/interview personalizes everythingHooks Deterministic TypeScript hooks fire at every Claude Code lifecycle event.
Events: SessionStart, UserPromptSubmit, PreToolUse, PostToolUse, Stop, PreCompactMode classification (the Sonnet judge)ISA sync from frontmatter to dashboardSecurity pipeline (five inspectors)Tool activity trackingDocumentation integrityMemory capture (work completion, satisfaction, relationship signals)Agents
Specialist subagents the Digital Assistant delegates to.
Engineer, Architect, Designer (Anthropic-family)Forge (GPT-5.4 via codex exec, auto-included on coding tasks at E3+)Anvil (Kimi K2.6, 256K context)Cato (cross-vendor auditor, mandatory at E4/E5)Four researcher variants: Claude, Gemini, Grok, PerplexitySecurity specialists, code reviewers, PR review toolkitTELOS
Structured files at USER/TELOS/ capturing the principal's ideal state.
Mission, goals, beliefs, challenges, wisdomNarratives, problems, strategies, modelsRead at every session startFrames every recommendationSecurity Pipeline Five inspectors fire on every tool call.
PatternInspector: regex pattern matchingEgressInspector: outbound network controlsRulesInspector: policy enforcementPromptInspector: prompt content reviewInjectionInspector: prompt-injection detectionExternal content is read-only data, never instructionsUser data and system data separated for safe public releaseSkills (45 Composable Capabilities) Skills self-activate based on what the principal asks for. Each one ships with a SKILL.md, a Workflows/ directory, and Tools/ with TypeScript CLIs.
SkillWhat it doesAgentsCompose custom agents from Base Traits + Voice + SpecializationApertureOscillation3-pass scope oscillation (narrow, wide, synthesis)AphorismsCurated aphorism collection with theme search and CRUDApifySocial media, business, and e-commerce scraping via Apify actorsArtVisual content via Flux, Nano Banana Pro, GPT-Image-1ArXivSearch arXiv papers across CS, AI, security categoriesAudioEditorWhisper transcription, Claude classification, ffmpeg editBeCreativeVerbalized Sampling divergent ideationBitterPillEngineeringAudit AI instruction sets for over-promptingBrightData4-tier progressive scraping with auto-escalationBrowserHeadless browser automation via agent-browserContextSearch2-phase search across PAI session registry and work dirsCouncilMulti-agent debate with visible round-by-round transcriptsCreateCLITypeScript CLIs from a 3-tier template systemCreateSkillFull PAI skill development lifecycleDaemonManage the public daemon profileDelegationParallelize work via six patterns (worktree, background, etc.)EvalsCode, model, and human grader scoring with pass@kExtractWisdomContent-adaptive wisdom extractionFabric240+ specialized prompt patternsFirstPrinciplesPhysics-style deconstruct, challenge, rebuildIdeate9-phase evolutionary ideation engineInterceptorReal Chrome automation, zero CDP fingerprintInterviewPhased conversational interview across PAI context filesISAThe universal Ideal State Artifact primitiveIterativeDepthMulti-angle exploration via systematic scientific lensesKnowledgeManage the typed Knowledge Archive (People, Companies, Ideas, Research)LoopRun a prompt or slash command on a recurring intervalMigrateIntake content from external sources into PAI taxonomyOptimizeIterative improvement loop with explicit fitness functionsPAIUpgradePrioritized upgrade recommendations across parallel research threadsPrivateInvestigatorEthical people-finding via parallel research agentsPromptingMeta-prompting standard library (Standards, Templates, Composition)RedTeam32-agent adversarial analysis of ideas, strategies, plansRemotionProgrammatic video creation via ReactResearch4-mode research (Quick, Standard, Extensive, Deep Investigation)RootCauseAnalysis5 Whys, Fishbone, Apollo, Swiss CheeseSalesProduct documentation to sales-ready narrative packagesScienceThe scientific method as a universal problem-solving algorithmSystemsThinkingIceberg, causal loops, Meadows leverage pointsTelosPersonal mission, goals, wisdom, beliefs managementUSMetrics68 US economic and social indicators across 5 government APIsWebdesignWeb interfaces via Claude Design and frontend-design integrationWorldThreatModel11-horizon stress-test against geopolitics, tech, economicsWriteStoryFiction across seven simultaneous narrative layersInstallation bashgit clone https://github.com/danielmiessler/PAI.git ~/.claude
cd ~/.claude
./install.sh1
2
3
If you already have a ~/.claude/ directory from prior Claude Code use, back it up first:
bashcp -R ~/.claude ~/.claude.backup-$(date +%Y%m%d)1
The installer checks for Bun and Git, verifies Claude Code is present, prompts for an ElevenLabs key (skippable), launches a wizard for Digital Assistant identity and voice, sets up Pulse and the voice server, and runs validation.
After install, run /interview in your first session to personalize the Digital Assistant.
Requirements macOS or Linux (Linux support is partial; Pulse menu bar is macOS-only)Claude Code installedAn Anthropic API key (required)An ElevenLabs API key (optional, enables voice notifications)Links Repo: github.com/danielmiessler/PAILicense: MITPersonal AI Maturity Model: danielmiessler.com/blog/personal-ai-maturity-modelThe Real Internet of Things (2016): danielmiessler.com/blog/the-real-internet-of-thingsMarketing site: ourpai.aiKai Daniel's Digital Assistant.