These records make the practice concrete: agentic systems, delivery infrastructure, and developer tools built around repeatability, governance, and accountable change.
Scope is intentionally narrow. Names, dates, classifications, descriptions, and links come from the public project catalog; no adoption, production, or outcome claims are implied.
Browse the complete ledgerRecent featured records from the catalog. Each entry names the mechanism the system is designed to make possible; the source link carries the implementation detail.
Record 01
Products
Git-native institutional memory for humans and agents: typed Markdown/YAML knowledge, graph retrieval over MCP, and a human-gated agent work loop. Open source and pre-launch; packages are not yet published.
Evidence basis: public repository and recorded project description no adoption or production outcome implied
Record 02
Infrastructure
A public agent delivery harness with durable work state, exact-candidate gates, content-addressed evidence, bounded agent-facing tools, deterministic checks, and an operator-owned release-authorization verdict. The documented local implementation is process-tested; live authenticated multi-provider delivery and external delivery remain unproven.
Evidence basis: public repository and recorded project description no adoption or production outcome implied
Record 03
Infrastructure
Define your agentic system once — rules, agents, pipelines, gates, knowledge — and compile it into native config for Claude Code, Codex, and GitHub Copilot. Plain files in git; a Rust CLI that never runs the agent loop.
Evidence basis: public repository and recorded project description no adoption or production outcome implied
Record 04
Plugins
Claude Code skills, compiled: a monorepo where plugins, marketplace, and catalog are generated from validated sources — quality rules, security audit, trigger evals, and a byte-exact CI drift gate.
Evidence basis: public repository and recorded project description no adoption or production outcome implied
Record 05
Infrastructure
A self-replicating Claude Code plugin marketplace, built as a software factory. The output is not plugins — it's the system that produces and governs plugins, live at registry.objectcore.ai.
Evidence basis: public repository and recorded project description no adoption or production outcome implied
The complete public catalog, ordered by its recorded date. Filter by system type without losing the source description or destination.
Showing 19 of 19
Products
Git-native institutional memory for humans and agents: typed Markdown/YAML knowledge, graph retrieval over MCP, and a human-gated agent work loop. Open source and pre-launch; packages are not yet published.
Infrastructure
A public agent delivery harness with durable work state, exact-candidate gates, content-addressed evidence, bounded agent-facing tools, deterministic checks, and an operator-owned release-authorization verdict. The documented local implementation is process-tested; live authenticated multi-provider delivery and external delivery remain unproven.
Infrastructure
Define your agentic system once — rules, agents, pipelines, gates, knowledge — and compile it into native config for Claude Code, Codex, and GitHub Copilot. Plain files in git; a Rust CLI that never runs the agent loop.
Plugins
Claude Code skills, compiled: a monorepo where plugins, marketplace, and catalog are generated from validated sources — quality rules, security audit, trigger evals, and a byte-exact CI drift gate.
Infrastructure
A self-replicating Claude Code plugin marketplace, built as a software factory. The output is not plugins — it's the system that produces and governs plugins, live at registry.objectcore.ai.
Agent systems
A durable pipeline of four Claude Managed Agents that turns a plain-language request into a validated, installable Claude Code plugin — designed, optimized, scaffolded, and independently reviewed behind a human gate.
Infrastructure
When an Azure DevOps work item goes Active, this service clones the repo, has Claude Code implement the change, and pushes a feature branch for review.
Plugins
Claude Code plugin for Sitecore, Umbraco, and Optimizely CMS development — analysis, scaffolding, and enhancement workflows.
Infrastructure
A fullstack TypeScript monorepo starter — Turborepo, Next.js, Bun, Drizzle + Postgres, Better Auth, Tailwind v4, and a full test/CI setup.
Plugins
Expert architect for designing and building Claude Code plugins. Covers the full extension ecosystem — plugins, skills, MCP servers, hooks, agents, and more.
Plugins
Maintenance and operations engine for Claude Code plugins — health scanning, dependency management, versioning, issue tracking, and runbooks.
Plugins
Go-to-market engine for Claude Code. Analyze your product, build a GTM strategy, generate launch content, and track your launch — all from the terminal.
Agent systems
Turns Claude Code into a hive mind — a coordinated colony of specialized agents that decompose, implement, verify, and ship features without manual intervention.
Plugins
Guides architects through Sitecore XP on AWS to Azure migrations — structured discovery, cross-reference analysis, phased estimation, and client-ready deliverables.
Plugins
Claude Code plugins that analyze enterprise CMS codebases, deliver actionable reports with severity ratings, and generate custom AI enhancements.
Agent systems
Claude Code plugin that transforms Claude into an autonomous coding agent with Plan → Execute → Review cycles and self-learning capabilities.
Agent systems
Multi-agent orchestration system for autonomous software development. Coordinates multiple AI agents to tackle complex engineering tasks.
Infrastructure
Enterprise MCP server that provides unified access to any CMS through a single, consistent interface.
Agent systems
A high-performance orchestration layer built to give "Industrial Clarity" and engineering precision to AI-assisted development.