Loading Lithora...
Please wait
Loading Lithora...
Please wait
What is Lithora?
Lithora is the work platform for engineering-led teams. Issues, docs, files and GitHub activity live in a single connected graph — and an AI agent reasons over that graph and acts on it, under your approval. This page is the long version: why it exists, what it does, and who it is for.
The problem
Modern work is spread across too many disconnected tools. Tasks in one app, documents in another, decisions in a third, files somewhere else. Every switch costs focus, context and momentum.
Five or more products for issues, docs, files, chat and delivery metrics. Information silos, and the context that explains a decision lives somewhere nobody thinks to look.
Every jump between apps costs refocus time. Across a week those minutes become hours of lost deep work, and the cost lands on the people doing the building.
When tools do not talk to each other, teams spend more energy chasing updates and re-typing the same status than on the work the status describes.
Lithora’s answer is one place where all of it connects, so the context that explains a piece of work is attached to the work itself.
Capabilities
Each one closes a workflow end to end. Everything else in the product exists to support these four.
Plan a sprint from a prompt, read the action plan, approve it, and the agent executes against your graph — under your permissions, with every run written to an audit trail. It also reviews pull requests and turns red builds into owned fix tasks.
Issues, docs, files, code and GitHub activity in a single graph, with semantic search and cited answers. Open a task and its files, discussion, pull requests and project are already there.
Event, condition, action — built in a visual editor and readable as configuration, with a run history you can debug. Per-team N8N workflows, on Scale, cover anything that lives outside Lithora.
Two-way sync. GitHub issues arrive as work items; work moves to done as pull requests merge and their issues close. Planning happens where the code already is.
Principles
Four commitments that decide how features behave when they conflict.
Tasks, files, notes and conversations are linked rather than filed in parallel silos. The connection is what makes the graph — and the agent reasoning over it — possible.
Task changes, notifications and whiteboard edits are broadcast to your team as they happen. No manual refresh, no wondering whether you are looking at yesterday.
The agent proposes and waits. Write actions are staged behind a confirmation gate, and they run as the person who approved them — never with elevated permissions.
Deleting a task or a project is a 30-day soft delete with a restore path, and the interface never reports a success the server has not confirmed.
Day to day
The things teams touch every day, including the ones that rarely make a feature list.
Overview, Board, Backlog, Timeline, Calendar, Tasks, Files and Integrations are sub-views of the same project. Board answers what is moving today; Timeline answers what ships this quarter and whether it is slipping.
The guest role is read-only and never billable — 5 free guests per team on Pro, 25 on Scale. Clients, execs, designers and contractors get access without a seat.
Mint a read-only public link showing project name, description and status plus task title, status, priority and due date. No member emails, no internal ids. Optional expiry, revocable at any time.
A new hire on your company domain can discover their colleagues’ team and request to join, instead of starting an empty workspace nobody else uses. Consumer email domains are excluded.
Mail that arrives through the inbound parser becomes a task, and the thread stays attached to it — so the customer message and the work it caused stay together.
A command palette (⌘K) and global search jump straight to a task, project, file or setting. Semantic search over the memory graph answers questions rather than returning a list.
Every seat gets personal storage and every team a shared pool, with files attached directly to tasks and projects. Add storage when you outgrow the plan.
Set a trigger, a condition and an action — when work moves to done, notify the channel — and read the execution log to see exactly what ran and when.
Under the hood
Six mechanics that explain most of the behaviour you will notice in the first week.
Everything you create — tasks, notes, files, whiteboard sketches — becomes a node in one graph. A task can reference a file, belong to a project, carry a discussion and link a pull request at the same time.
Changes propagate as they happen. Presence shows who is in a project, and live updates keep the board honest without anyone refreshing it.
The command palette handles navigation; the memory graph handles meaning. Ask who worked on the checkout flow and the answer arrives with the tasks, PRs and docs it came from.
Ask from inside a project and the answer is scoped to that project. Request a breakdown and it uses your real backlog, not a generic template.
Roles decide who administers a team, manages billing or changes settings. Personal storage stays private, team storage is shared, and every query is scoped to its tenant.
Start from one of over 300 project templates, or save your own setup and reuse it the next time the same shape of work comes around.
Who it is for
Two paid plans, monthly billing, a 7-day trial on both. There is no free tier and no annual contract — we do not sell one.
Keep tasks, files and notes in one private workspace. Pro starts at $10 per seat, per month, with a 7-day trial and 5 GB of personal storage per seat.
Move without stitching five SaaS tools together. Pro covers up to 5 teams of 20, with automations, GitHub-native sync, whiteboards and the agent from day one.
Scale is $29 per seat, per month: up to 15 teams of 50, 20 GB personal per seat, 400 GB shared, 1 GB file uploads, audit logs, advanced analytics and SSO with SCIM provisioning.
Larger organisations with procurement, governance or contractual requirements can talk to us about a custom Enterprise agreement built on Scale’s entitlements.
Trust
Lithora holds your issues, your docs and a live connection to your code host, so the honest version matters more than the impressive one.
Read the security postureHttpOnly cookie sessions, Origin-checked mutations, fail-closed webhook signature verification, SSRF-guarded outbound calls, and per-tenant scoping at the data layer.
No success message until the server confirms it. No optimistic fakes, including in the agent’s own reporting.
Export and import whenever you want, and AI features never train on your content. Deleted work is restorable for 30 days before it is purged.
Lithora is not SOC 2 or ISO certified, and we will keep saying so until an audit is complete. Status is public and the roadmap carries ship states, not dates.
Start a 7-day trial on Pro or Scale, or open the demo first — it runs on sample data and needs no signup.
Questions first? Talk to our team.