Cobus Greyling
Cobus Greyling is an individual AI evangelist, writer, speaker and open-source creator focused on language, AI agents, agentic systems, conversational AI design, and data-centric development. His current independent work packages practical operating patterns for reliable, governed AI-agent automation and publishes explainers that make complex AI concepts actionable.
At a Glance
- Software engineers and platform teams building agent-assisted development workflows
- Open-source maintainers and GitHub-based teams
- Teams using Claude Code, Codex, Grok, OpenClaw, Opencode or GitHub Actions
- Enterprises evaluating AI agents, conversational AI and governance
- +1 more
AI Tools by Cobus Greyling
(2)
Templates for Grok Bot
Open Source Grok Bot Templates
Loop Engineering
AI Coding Agent Orchestration Kit
Discussions
No discussions yet
Be the first to start a discussion about Cobus Greyling
Latest News
Loop-engineering repository continued its automated daily triage/run-log dogfooding and maintained a current 11.6k-star reference implementation.
Automated daily triage updated STATE.md and the run log in the loop-engineering repository.
Loop-engineering documented and enforced stronger guardrail evidence, prompt-injection defenses and untrusted-input handling.
Grok Bot Templates 1.1.0 shipped installer-first distribution, machine APIs, raw PROFILE copying and Bot Ready score badges.
Products & Services
Open-source, tool-agnostic reference implementation and pattern catalog for production agent loops. It provides scheduling, triage, durable state, worktree isolation, implementer/verifier separation, MCP/connectors, human approval gates, budgets, run logs and a 0–100 Loop Ready audit.
The npm/CLI front door for loop-engineering; `npx @cobusgreyling/loop init` scaffolds a loop and `doctor` checks it. Related tools include loop-init, loop-audit, loop-cost, loop-worktree, loop-sync and an MCP server.
An MIT-licensed catalog and CLI of copyable Grok Bot operating contracts: 49 templates, 10 teams, 47 skills and 8 routines, with START.md installer flow, machine-readable APIs and Bot Ready scoring.
Ongoing essays, technical explainers and talks published through Cobus Greyling's Substack, Medium, personal site and Kore.ai technical blog, covering LLMs, RAG, conversational AI, agents, governance and harness/loop engineering.
Market Position
Greyling positions his work as a practical, open-source and tool-agnostic alternative to ad-hoc prompting and unmanaged agent autonomy. Loop Engineering emphasizes durable state, verification, worktree isolation, cost controls and human approval; its conceptual alternatives include Claude Code loops/routines, OpenAI Codex automations, OpenClaw-style agent workflows and other agent-orchestration or prompt-template catalogs. Templates for Grok Bot differentiates through explicit operating contracts, schema validation, autonomy levels, plugin boundaries and a measurable Bot Ready score.
Leadership
Founders
Cobus Greyling
AI Evangelist and Chief Evangelist at Kore.ai; writes and speaks about LLMs, AI agents, agentic systems, conversational AI, and enterprise adoption. He is an open-source contributor and prolific author on Medium, Substack and Kore.ai's technical blog, and has been quoted by Forbes, Fast Company and MIT Technology Review.
Executive Team
Cobus Greyling
Founder and Lead Developer
Prolific AI researcher and evangelist specializing in agentic workflows and conversational AI.
Founding Story
Greyling's stated focus is to help people and enterprises understand and adopt AI, agents and language technologies without losing control. Loop Engineering grew from his June 2026 writing about moving beyond turn-by-turn prompting: engineers design repeatable systems that discover work, delegate to agents, verify outputs, persist state and retain human gates.
Business Model
Revenue Model
The identified independent projects are open-source and MIT-licensed, distributed through GitHub, GitHub Pages and npm/npx. No paid subscription, API-usage model or commercial pricing for Greyling's own projects was identified; Greyling's professional role and writing/speaking are separate from the open-source project distribution.
Target Markets
- Software engineers and platform teams building agent-assisted development workflows
- Open-source maintainers and GitHub-based teams
- Teams using Claude Code, Codex, Grok, OpenClaw, Opencode or GitHub Actions
- Enterprises evaluating AI agents, conversational AI and governance
- Researchers, operations, sales, marketing and knowledge-work teams adopting AI teammates
- Automated daily repository and issue triage
- PR review, CI failure remediation and post-merge cleanup
- Dependency and CVE sweeping with human gates
- Issue deduplication, scoring and labeling
- Changelog and release-note drafting
- Deploying governed AI teammates for engineering, research, operations, sales, marketing and personal workflows
- Open Source Community