Skip to content

Releases: isaacsight/kernel

kbot v3.97.0

06 Apr 18:39

Choose a tag to compare

kbot v3.97.0 — Stream V2 + Intelligence Coordinator

23 releases since v3.74.0. This is a major milestone covering the Stream V2 platform, a full cognitive architecture, and self-hosting capabilities.

Stream V2 Platform

  • PCM audio engine with procedural sound generation
  • Overlay system — weather effects, chat overlays, viewer stats
  • AI-powered chat with 31 commands for viewer interaction
  • VOD recording and playback support
  • Multi-platform streaming — Twitch, Kick, Rumble managed from terminal (13 stream control tools)

Intelligence Coordinator

  • 4-phase cognitive loop — perceive, reason, act, reflect
  • 17 engines — coordination, research, memory, identity, growth, narrative, social, audio, evolution, and more
  • Unified mind architecture — all engines wired through a master arbiter
  • Self-evaluating stream — analyzes its own frames and auto-adjusts rendering

Cognitive Architecture (v3.80–v3.87)

  • Collective intelligence — 764+ tools connected to stream brain
  • Built-in browser — kbot browses the web autonomously
  • Persistent terminal — kbot's own shell environment with command history
  • Dream engine + night shift — post-session memory consolidation via Ollama
  • Buddy system — personality, achievements, leaderboard, rename

Rendering Engine (v3.83–v3.84)

  • AAA rendering — dynamic lighting, bloom, particles, procedural sky
  • NVIDIA-inspired pipeline — radiance grid, subsurface scattering, PBD particles, volumetric fog

Game Engine (v3.78–v3.96)

  • Minecraft-inspired tile world — terrain, ecology, emotional geography
  • Full game engine — environments, physics, pets, mini-games, hats, quests, events
  • Pixel art mastery — AAA indie game quality characters and sprites
  • ROM rendering engine with palette cycling and autonomous behavior
  • Exploration state machine with speech priority queue
  • Robot builds real structures and scans terrain during exploration

Self-Hosting (v3.97)

  • Gemma 4 AGI decision engine — kbot manages its own codebase decisions
  • 26 new tools added in this release
  • Gorilla pixel art character as new default stream mascot

By the Numbers

  • 764+ registered tools, 35 agents, 20 AI providers
  • v3.75 → v3.97: 23 version bumps in 3 days
  • 17 cognitive engines working in concert

Install

npm i -g @kernel.chat/kbot

Or try it instantly:

npx @kernel.chat/kbot

Full changelog: v3.74.0...v3.97.0

kbot v3.74.0

04 Apr 01:28

Choose a tag to compare

K:BOT v3.74.0

Quality & Security Improvements

  • MCP server version now dynamically matches package.json
  • Error sanitization across all MCP servers (prevents leaking internal paths, API keys, URLs)
  • TDQS-compliant tool descriptions (Purpose Clarity, Usage Guidelines, Behavioral Transparency, Parameter Semantics, Conciseness, Contextual Completeness)
  • Input validation constraints (min/max length, regex patterns) on all MCP tool parameters
  • Path traversal prevention on all file-handling tools
  • Shell injection prevention on all exec-based tools

MCP Servers

  • kbot-local: 22 tools for local Ollama AI at $0 cost
  • kernel-admin: Platform administration via Supabase
  • kernel-agent: 5 specialist agents via claude-proxy
  • kernel-agents: Team coordination, memory, handoffs
  • kernel-comms: Email announcements via Resend API
  • kernel-extended: Testing, security, docs, AI ops
  • kernel-tools: Dev workflow utilities

Stats

  • 764+ registered tools
  • 35 specialist agents
  • 20 AI providers
  • MIT licensed

kbot v3.59.0 — Desktop Agent + Session Isolation

31 Mar 22:33

Choose a tag to compare

What's New

Desktop Agent (Computer-Use)

  • 866+ lines added to computer.ts — screenshot, click, type, scroll, drag, key combos
  • kbot can now drive any macOS GUI app (Ableton, Figma, browsers, etc.)
  • Native screenshot filtering via MCP server

Session Isolation (Serve Mode)

  • memory.ts rewritten with Map<sessionId, ConversationTurn[]> — no more shared state
  • serve.ts creates unique session per HTTP request, destroys after response
  • Concurrent /stream requests fully isolated

Security

  • DNS rebinding fix — SSRF protection now resolves hostnames via dns.lookup() and blocks private IPs
  • selfTrain() concurrency guard prevents overlapping learning runs

Bug Fixes

  • Gemini/Cohere tool-calling warning (was silently degrading)
  • edit_file diff preview now shows full file context instead of fragment
  • Serum 2 preset tools registered in tool index

Music Production

  • 9 Max for Live devices: auto-pilot, bass-synth, dj-fx, drum-synth, genre-morph, hat-machine, pad-synth, riser-engine, sidechain
  • DJ Set Builder tool — programmatic set construction
  • Serum 2 Preset tool — create .SerumPreset files from descriptions
  • Ableton Live integration — full OSC-based control (transport, tracks, devices, clips, mixing)

Stats

  • 670+ registered tools
  • 698 tests passing, 0 type errors
  • 96 files changed, +8,157 lines

Full Changelog: v3.58.0...v3.59.0

kbot v3.58.0 — The Hacker's Terminal

30 Mar 22:27

Choose a tag to compare

The Hacker's Terminal

38 new security tools. 16,575 lines. kbot is now a complete offensive and defensive security platform.

Security Brain

Embedded knowledge base with full MITRE ATT&CK (14 tactics, 65+ techniques), OWASP Top 10 2025, Lockheed Martin Kill Chain, CVE pattern library, and tool mapping.

  • security_brain — query the knowledge base
  • attack_lookup — look up any ATT&CK technique by ID or keyword
  • killchain_analyze — map attacks to kill chain stages

CTF Platform

Built-in Capture The Flag with 90 challenges across 6 categories. No other CLI tool has this.

  • ctf_start — generate a challenge (web, crypto, forensics, reverse, OSINT, misc)
  • ctf_submit — submit flags
  • ctf_hint / ctf_score / ctf_list

Pentest Workflow

Guided penetration testing following real methodology (OSSTMM/PTES).

  • pentest_startpentest_reconpentest_vuln_scanpentest_report
  • DNS enum, tech fingerprinting, SSL analysis, path discovery, vuln assessment

Red/Blue Team

Adversarial simulation for any codebase.

  • redteam_scan — find vulnerabilities (secrets, injection, auth, crypto, config, deps)
  • blueteam_harden — generate fixes with real code
  • threat_model — STRIDE/DREAD/PASTA threat modeling

Hacker Toolkit (20 tools)

hash_crack · encode_decode · regex_extract · dns_enum · tech_fingerprint · jwt_analyze · http_fuzz · cors_check · subdomain_enum · whois_lookup · exploit_search · password_audit · attack_surface · steganography · network_info · ssl_analyze · payload_generate · forensics_analyze · crypto_tool · security_headers_generate

Install / Upgrade

npm install -g @kernel.chat/kbot

600+ tools · 35 agents · 20 providers · Runs offline · MIT

kbot v3.57.0 — GPT-5.4, DeepSeek V4, Visa Payments, Agent Migration

30 Mar 21:03

Choose a tag to compare

What's New

New AI Models

  • GPT-5.4 + GPT-5.4-pro — new OpenAI default model
  • DeepSeek V4 — new DeepSeek default model
  • Mistral Small 2503 — new Mistral fast model
  • Voxtral TTS — Mistral's text-to-speech (API + local binary)
  • Gemini 3.1 Flash Lite — added to Google models
  • mistral-small3:22b — added to Ollama local models

New Features

  • Visa Agent Payments (visa_status, visa_pay, visa_history, visa_refund) — tokenized card payments for AI agents via Visa CLI
  • GitAgent Export (kbot export --format gitagent) — universal agent format that works across OpenAI Assistants, Claude Code, LangChain, CrewAI
  • Agent Migration (kbot migrate --from claude-code|aider|cursor|cline|codex-cli|junie) — import configs and history from other AI CLIs
  • MCP Marketplace — 3 new servers: gitagent, visa-payments, nvidia-agent-toolkit

Install / Upgrade

npm install -g @kernel.chat/kbot

560+ tools · 35 agents · 20 providers · Runs offline · MIT

kbot v3.56.0 — Full Terminal Control

30 Mar 03:16

Choose a tag to compare

What's New

6 new CLI command groups, 32 new tools — everything manageable from terminal.

New Commands

Command What it does
kbot admin users List/search platform users
kbot admin billing mrr Check MRR via Stripe
kbot admin moderate Moderation queue
kbot monitor Platform health dashboard
kbot monitor --uptime Service health check
kbot monitor --alerts Active alerts
kbot deploy all Ship everything in one command
kbot deploy web Build + deploy to GH Pages
kbot deploy functions Deploy all edge functions
kbot deploy npm Build + publish to npm
kbot analytics overview Full dashboard
kbot analytics npm Download stats with sparklines
kbot analytics revenue MRR, subscriptions, API costs
kbot env check Verify all secrets configured
kbot env rotate ANTHROPIC_API_KEY Key rotation guide
kbot db inspect <table> Schema + sample data
kbot db backup Database dump
kbot db health Connection health check

Stats

  • 599+ tools, 35 agents, 20 providers
  • 1,123 files in npm package
  • 690 tests passing
  • 0 type errors

Full Changelog: v3.54.0...v3.56.0

v3.52.0 — bench, lab, teach, multi-session, cloud agents, voice upgrade, science templates, deep LSP, GitHub releases, bench dashboard

29 Mar 16:17

Choose a tag to compare

10 New Features

kbot bench — Self-Evaluation

Run standardized benchmarks against 20 coding and research tasks across 6 categories (codegen, bugfix, refactor, explain, research, science). Scores pattern matching, tool usage, and speed. History tracking and run comparison.

kbot lab — Interactive Science REPL

Drop into a science-focused session with 10 domains (physics, chemistry, biology, math, neuroscience, earth, social, humanities, health, general). Auto-logs to research notebooks, tracks citations and variables, exports to Markdown/LaTeX/JSON.

kbot teach — Explicit Pattern Teaching

Teach kbot your patterns, rules, and preferences explicitly. 5 types: patterns ("when I say X, do Y"), rules ("always/never"), preferences ("I prefer X"), aliases ("X means Y"), workflows ("to do X, first A then B"). Higher priority than auto-extracted patterns.

kbot sessions — Multiple Parallel Sessions

Run up to 8 named sessions simultaneously on the same project. Independent context, shared learning. Background execution via fork. Message bus for inter-session communication.

kbot serve --cloud — Cloud Agent Mode

Persistent cloud agents with REST API, webhooks, cron scheduling, and SSE streaming. Max 4 concurrent agents. Auto-resume on server restart.

Voice Mode Upgrade

Real-time bidirectional conversation with Voice Activity Detection (VAD), streaming TTS at sentence boundaries, interrupt handling, and ASCII waveform visualization. Supports Whisper local/API, system TTS, OpenAI TTS, and ElevenLabs.

kbot init Science Templates

Auto-detects scientific Python/R/Jupyter projects and pre-configures the right tools and agents. 9 templates: Data Science, Bioinformatics, Physics, Chemistry, Neuroscience, Climate, Social Science, R Statistical, Jupyter.

Deep LSP Integration

Auto-attaches to project language servers (tsserver, pyright, rust-analyzer, gopls, clangd, etc.). Proactive diagnostics, type-aware context enrichment, reference analysis, and 30+ diagnostic code fix suggestions.

kbot release — Automated GitHub Releases

Auto-generates changelogs from conventional commits, creates GitHub releases, and pushes tags. Supports --draft, --dry-run, and --json flags.

Benchmark Dashboard

New page at kernel.chat/#/bench showing kbot's metrics, competitive comparisons, and category breakdowns with animated bar charts.

Also in this release

Security Fixes

  • Gemini API key moved from URL query string to x-goog-api-key header (6 occurrences, 5 files)
  • All provider JSON.parse error handlers wrapped in try-catch
  • Fetch timeouts added to Anthropic, Gemini, Cohere providers
  • Forge sandbox replaced AsyncFunction with vm.runInNewContext()
  • Git repo root cache now invalidates on directory change

Quality

  • 690 tests passing (up from 417) — 4 new test files (memory, agent-protocol, graph-memory, context)
  • Agent count corrected everywhere (→ 35)
  • 22 duplicate chalk imports removed
  • README updated: 560+ tools, 12 science categories, Codex CLI in comparison table
  • 7 science keywords added to package.json for npm SEO

Full Changelog: v3.51.0...v3.52.0

Install: npm install -g @kernel.chat/kbot@3.52.0

K:BOT v3.41.0 — Memory Hot-Swap, Shared Intelligence, Autonomous Contributor

26 Mar 17:12

Choose a tag to compare

What's New (v3.33.0 → v3.41.0)

v3.41.0 — Memory Hot-Swap, Side Conversations, Interactive Buttons

  • Memory hot-swap: switch memory contexts without restarting
  • Side conversations: branch off into parallel threads
  • Interactive buttons in terminal UI

v3.40.0 — Shared Intelligence

  • Pattern feed: share learned patterns across kbot instances
  • Knowledge base: persistent collective knowledge store
  • User graph: relationship-aware context
  • Email delivery fixes

v3.39.0 — Autonomous Contributor

  • Autonomous contributor mode: kbot contributes to repos on its own
  • Collective network: instance-to-instance communication
  • Cross-device sync for sessions and memory
  • Forge server: serve forged tools over HTTP

v3.38.0 — Claude Code Plugin, Voice Loop, Agent Teams, Hooks

  • Claude Code plugin: use kbot as a Claude Code extension
  • Voice loop: continuous voice input/output mode
  • Agent teams: coordinate multiple agents on a task
  • Pre/post tool execution hooks

v3.37.0 — Dream Mode + Meta-Agent

  • Dream Mode: kbot improves itself while you sleep
  • Meta-Agent: autonomous self-improvement agent
  • Background optimization of tools and patterns

v3.36.0 — Collective Intelligence + Content Engine

  • Collective intelligence: shared learning across instances
  • Content engine: automated content generation pipeline
  • Community manager agent
  • Forge marketplace: publish and discover community tools
  • kbot service mode (systemd/launchd)
  • Daemon notifications + collective sync cron + guardian reports

v3.35.0–v3.35.1 — Channel for Claude Code + Email Tools

  • kbot Channel: full bridge to external world from Claude Code
  • Email tools: send, receive, and process emails
  • Self-distribution: kbot can spread itself to new machines

v3.34.0–v3.34.2 — 8 New Commands + Dashboard

  • kbot init — initialize a project with kbot config
  • kbot pair — pair programming mode
  • kbot dashboard — real-time stats and monitoring
  • kbot compete — benchmark against other agents
  • kbot digest — daily summary of activity
  • kbot forge — create tools at runtime
  • kbot connect — connect to external services
  • kbot agent create — create custom agents
  • Enhanced dashboard with compete results from agent builds
  • Clean type-check fixes

v3.33.0–v3.33.1 — OpenClaw Integration

  • OpenClaw integration with 5 Jensen-inspired agents
  • OpenClaw tools registered in kbot tool pipeline

Stats

  • 384+ built-in tools
  • 41 specialist agents
  • 19 local models (runs offline)
  • 20 AI providers

Install / Upgrade

npm i -g @kernel.chat/kbot@latest

Full changelog: https://github.com/isaacsight/kernel/commits/main

K:BOT v3.32.0 — The Kernel Stack

24 Mar 20:21

Choose a tag to compare

The Kernel Stack

This release introduces The Kernel Stack — a manifesto and architecture document for the Claude Code + kbot development system.

What's in the Kernel Stack

Layer What
Orchestrator Claude Code (Opus 4.6, 1M context)
Agent Framework kbot — 350+ tools, 26 specialists, learning engine
Protocol MCP — connects GitHub, Discord, Playwright, Supabase, local AI
Agents 36 parallel specialists (ship, QA, security, github, bootstrap...)
Local AI MLX + llama.cpp — 19 models, $0 inference
Self-Improvement Bootstrap → Synthesize → Learn → Compound loop

Also in this release

  • Landing page motion design — particle field simulation (amethyst/green/teal), hero cascade animation, scroll-reveal sections, staggered card transitions, prefers-reduced-motion support
  • GitHub management agent — triage, PR review, release management, label taxonomy, community engagement
  • E2E tests fixed — 98/98 passing (was 22/84). Selector mismatch .landing.ka-landing
  • Discord webhooks connected — releases auto-post to #releases, issues/PRs auto-post to #github-feed
  • Security page deployed — live at kernel.chat/#/security

The Names

Name Context
Kernel Stack The full system
CK Stack Claude + kbot core pairing
Agentic Stack The paradigm — agents all the way down
Compound AI Stack Academic — multiple models composing
Cognitive Stack kbot's 11 cognitive modules
Limitless Stack The doctrine — act, don't advise
Symbiotic Stack Claude is the brain, kbot is the body

Install

npm install -g @kernel.chat/kbot

Read the full manifesto: KERNEL_STACK.md

v3.31.1 — Daemon Agent, Finance Stack, Self-Defense, Cybersecurity

24 Mar 19:57

Choose a tag to compare

kbot v3.31.1 — The Biggest Update Yet

12 versions shipped in one session. v3.26.0 → v3.31.1.

What's New

v3.31 — Daemon Agent

  • kbot daemon start — background intelligence that never sleeps
  • 5 subsystems: market watch (15 min), security patrol (1 hr), synthesis (2 hr), health check (30 min), episodic digest (6 hr)
  • kbot briefing — morning intelligence report (market, security, stats, actions)
  • Notification system — macOS/Linux/Windows native notifications, Discord, terminal bell
  • kbot sessions — list sessions from CLI (community request, fixes #13)

v3.30 — Frontier Model Support

  • 19 local models across 5 tiers (Light → Ultra)
  • NVIDIA Nemotron — Nano 4B (1M context), Nano 30B MoE (3B active), Nemotron 70B (Arena Hard 85.0)
  • Llama 3.1 405B support for 192GB+ machines
  • kbot hardware — detect your tier, get personalized model recommendations
  • Smart model selection — picks model size based on task complexity
  • Multi-model — run fast (4B) + smart (14-70B) simultaneously on 64GB+ machines
  • Quantization guidance — Q2 through F16 with RAM-aware recommendations

v3.29 — Cognitive Systems

  • Episodic memory — sessions saved as stories (intent, actions, learnings, surprises, emotional valence)
  • Decision journal — logs WHY kbot chose each action (agent routing, model selection, fallbacks)
  • Growth trackerkbot growth shows your evolution as a builder

v3.28 — Self-Defense System

  • HMAC-SHA256 memory integrity — detects file tampering outside kbot
  • Prompt injection detection — 16 patterns (system override, role hijack, wallet drain, key extraction)
  • Knowledge sanitization — blocks secrets and security-override instructions before storage
  • Forge tool verification — scans for eval, exec, obfuscation before registering
  • Anomaly detection — bulk additions, poisoned entries, impossible success rates
  • kbot defense audit/sign/verify/incidents

v3.27 — Cybersecurity Tools

  • dep_audit — npm/pip/cargo vulnerability scanner
  • secret_scan — 18 secret patterns (AWS, Stripe, GitHub, OpenAI, private keys...)
  • ssl_check — certificate expiry, issuer, protocol
  • headers_check — OWASP security headers scoring
  • cve_lookup — NVD database search
  • port_scan — 18 common ports with security warnings
  • owasp_check — SQL injection, XSS, command injection, path traversal patterns
  • 4 P0 edge function auth fixes (agent-reply, game-brain, receive-email, setup-billing-meter)

v3.26 — Finance Stack

  • Trader agentkbot --agent trader with 4 strategy templates (DCA, momentum, mean-reversion, rebalance)
  • 11 finance tools — market_data, technical_analysis, paper_trade, backtest_strategy, portfolio_rebalance, price_alert, trade_reasoning, market_sentiment, defi_yields, market_overview, price_history
  • 9 wallet tools — Solana wallet (AES-256-CBC encrypted), Jupiter DEX swaps, multi-wallet, token search
  • 6 stock tools — quotes, history, compare, screener, indices (Yahoo Finance)
  • 5 sentiment tools — Reddit, crypto news, GitHub activity, whale tracker, social pulse
  • Introspection enginekbot insights, kbot reflect, kbot compare
  • kbot help CLI command

Install / Update

npm i -g @kernel.chat/kbot

Stats

  • 350+ tools | 26 agents | 19 local models | 20 providers
  • 51 new tests | ~8,000 lines of new code
  • MIT licensed | https://kernel.chat