Lore

Coding agent harness.

Lore wraps your coding agent in a harness — persistent memory, rule enforcement, orchestrated delegation, and work tracking. Your agent starts every session knowing your project instead of starting over. Plain files, git-versioned, zero runtime dependencies.

Quick Start

npx create-lore my-project
cd my-project

Then open a session in your editor:

No configuration needed. Your first session gets a full context banner immediately.

What You Get

Sessions accelerate instead of resetting. The harness loads your project identity, rules, active work, available agents, and a map of everything your agent knows at session start. No re-explaining.

Semantic search and a live docs UI — highly recommended. Tell your agent to start the docs sidecar — it pulls the Docker image, configures ports, and launches semantic search over the full knowledge base plus a live MkDocs site. Without Docker, agents fall back to Grep/Glob search.

Snags, gotchas, quirks become fieldnotes that persist. When your agent hits an API quirk, an encoding edge case, or a deployment snag, the harness captures that as a fieldnote. That fieldnote loads in every future session. The mistake happens once, the fix persists.

Rules are enforced, not just documented. Your coding standards, docs rules, and security policies are injected before every file write. The agent sees the relevant rules right when it matters.

One knowledge base, every platform, every repo. Capture a fieldnote in Claude Code — it's available in Cursor and OpenCode. Link repos to one hub — they all share the same knowledge. No copying, no drift. See it in action.

Complex work delegates to focused workers. The harness orchestrates delegation: when work benefits from a fresh context window, it spawns workers loaded with curated skills and rules. Compound tasks split across parallel workers for maximum throughput.

Before / After

Without a harness — Every session starts cold. You re-explain your project, the agent re-discovers API quirks, makes the same mistakes, and yesterday's debugging session is gone.

With Lore — The agent knows your project. Fieldnotes from last week load automatically. Rules are enforced at write-time. Active roadmaps surface at startup. Complex work delegates to focused workers.

How It Works

Lore is a harness built from markdown files, hooks that shape agent behavior, and scripts that keep everything consistent.

• Fieldnotes (.lore/fieldnotes/) — Snags and patterns captured from real work. Available on-demand in every session.
• Skills (.lore/skills/) — Procedural capabilities and reusable commands.
• Agents (.lore/agents/) — Worker agents for delegated tasks, loaded with relevant skills and fieldnotes per-task.
• Docs (docs/) — Project context, rules, environment knowledge, runbooks, and work tracking.
• Hooks (.lore/harness/hooks/, .cursor/hooks/, .opencode/plugins/) — Inject context at session start, enforce rules before writes, nudge knowledge capture during work.
• Scripts (.lore/harness/scripts/) — Platform sync, validation, nav building.

All hooks are plain JavaScript you can read in minutes. They don't make network requests, execute shell commands, or access anything outside your project directory.

Working Across Repos

Tell your agent to link work repos to the hub — it generates configs so hooks fire from the hub even when you open the work repo directly. One hub, many repos, shared knowledge. See the cross-repo guide.

Commands

Platforms

All platforms share the same knowledge base. Skills, fieldnotes, agents, and rules written once sync to platform-specific formats automatically.

Documentation

Full docs: lorehq.github.io/lore-docs

Contributing

See CONTRIBUTING.md for development setup and guidelines.

Security issues: see SECURITY.md.

License

Apache-2.0