agilerr.app
Agile delivery without the hosted stack.
Agilerr is a local-first scrum board with a Go API, PocketBase auth and storage, fast backlog and kanban views, AI-assisted planning, and a clean release path from single binary to Docker Compose.
Plan across the full hierarchy
Projects, epics, features, stories, tasks, and a dedicated bug workflow stay cleanly separated.
Local-first by default
Run from a single binary or Docker Compose. PocketBase handles auth and storage without external services.
AI Add when you want it
OpenAI-powered planning helps flesh out epics, features, stories, and bugs while keeping developer tasks manual.
Permission-aware teams
Per-project memberships and role-driven actions keep editing, deleting, AI access, and admin workflows controlled.
Downloads
One-click startup, whichever path you prefer.
Ship a single binary to a laptop, or pull a Docker image and keep the instance inside Compose. Both paths keep the same product and the same data model.
Linux x64
https://agilerr.app/downloads/agilerr-linux-amd64.tar.gz
macOS Apple Silicon
https://agilerr.app/downloads/agilerr-darwin-arm64.zip
macOS Intel
https://agilerr.app/downloads/agilerr-darwin-amd64.zip
Windows x64
https://agilerr.app/downloads/agilerr-windows-amd64.zip
Docker Compose manifest
https://agilerr.app/install/docker-compose.yml
Quick start
Get running in minutes.
Agilerr is designed to be dropped onto a machine and started without a control plane. These are the paths most teams will care about first.
Run the binary locally
curl -L https://agilerr.app/downloads/agilerr-linux-amd64.tar.gz | tar xz
./agilerrThe binary starts PocketBase, the Go API, and the embedded frontend in one process.
Use Docker Compose
curl -O https://agilerr.app/install/docker-compose.yml
docker compose up -dIdeal for a team sandbox, homelab instance, or keeping your data under a mounted volume.
Configure AI planning
export OPENAI_API_KEY="sk-..."
export OPENAI_BASE_URL="https://api.openai.com"
./agilerrAI Add stays visible but disabled until an OpenAI key is configured.
Inside the product
Context-first kanban
Move through the hierarchy one layer at a time. Boards stay readable because only the direct child type is visible for the current route.
Structured backlog filters
Filter by type and tags, collapse noisy descriptions, and preserve just enough parent context when intermediate levels are skipped.
FAQ
Everything people ask before rollout.
These are the operational details that matter when teams evaluate a self-hosted planning tool: environment variables, AI behavior, permissions, and release flow.
Which environment variables matter on day one?
For a local trial, none are strictly required. Agilerr can generate a local port and API key automatically. For a stable install, set ADMIN_EMAIL, ADMIN_PASSWORD, HTTP_ADDR, PB_DATA_DIR, and AGILERR_API_KEY. Add OPENAI_API_KEY and optionally OPENAI_BASE_URL plus OPENAI_MODEL when you want AI Add enabled.
How does AI Add work?
AI Add opens a compact planning conversation tied to the current project or parent item. The assistant asks one question at a time, proposes sibling items for review, and lets you accept, reject, or edit each proposal before anything is created. Conversations are summarized and stored for reuse.
Can I run Agilerr without sending data to OpenAI?
Yes. Without OPENAI_API_KEY, the product still supports backlog management, kanban, bugs, permissions, API access, MCP, and all core workflows. The AI Add controls remain visible but disabled so the capability is discoverable without being active.
What happens when upgrades need migrations?
The binary now stamps its own version and stores the database schema version in PocketBase metadata. Future releases can ship ordered migrations that run automatically when the database version is behind the binary.
Where do I start after login?
Start on the project dashboard. It gives quick links into kanban, backlog, bugs, API docs, and the assigned-to-you summary. From there, drill into the hierarchy route by route.
How do I manage the backlog without losing the thread?
The backlog keeps the hierarchy visible while letting you filter by type and tags. Descriptions collapse by default, so large projects stay readable even with a deep structure.
Can bugs stay separate from the core hierarchy?
Yes. Bugs have their own page, their own board, and start in triage. They do not sit inside the epic-feature-story-task chain.
What can admins control?
System admins can manage users, temporary passwords, project memberships, API docs, and MCP docs. Project settings manage project metadata and the color language used for item types and statuses.
Support Agilerr
If this saves your team time, back the next release.
Donations help fund release packaging, migration tooling, docs, and AI workflow polish. Keep the landing page links as stubs for now and wire them to your preferred funding platform later.