AI coding agents are fast. Keeping track of their work is not.
Continuity turns founder input, agent runs, code changes, blockers, decisions, and proof into a shared work graph — so you know what changed, what’s blocked, and what can safely run next.
Start with one workflow
Continuity works alongside Codex, Claude Code, Cursor, and custom runners. Agents pull current work context, execute a scoped task, then report back with changes, evidence, blockers, and recommended next steps.
Install the CLI, connect your workspace token, and verify that agents can read and update shared work state.
npm install -g @ligence/continuity
continuity login --api-token <workspace-token>
continuity setup --json
continuity doctor --jsonPaste a short onboarding brief into your coding agent so it can install Continuity, verify access, pull its first work item, and report progress back.
Copies Markdown for Codex, Claude Code, Cursor, or another runner.
# Install Continuity
Install the hosted Continuity CLI and connect this workspace.
1. Install the CLI:
npm install -g @ligence/continuity
2. Ask the user for the workspace token if it has not been saved yet.
3. Save the token:
continuity login --api-token <workspace-token>
4. Run setup and verification:
continuity setup --json
continuity doctor --json
5. Pull your first work packet:
continuity next --agent codex --horizon overnight --compact --json
6. Before doing work, inspect the assigned node:
continuity agent-context --node <node-id> --agent codex --json
7. After work, reconcile the run:
continuity report-delta --node <node-id> --kind completed --summary "<what changed>" --evidence "<proof>" --json
Rules:
- Do not ask for CONTINUITY_WORKSPACE_ID.
- If you are not Codex, replace codex in --agent flags with your agent name.
- Treat the graph as the source of truth for goals, blockers, evidence, and next action.The execution graph for AI-assisted teams
An unguided agent drifts across competing signals until a human redirects it.
The agent starts from a prompt, follows several fading traces toward side quests and stale context, then needs a human redirect before reaching partial output.
prompt agent side quest stale thread unfiled debt partial output human redirectEach agent starts from a prompt, transcript, or ticket. Context gets stale, discoveries disappear, and humans have to reconstruct what is safe to do next.
A work graph connects the active goal to tasks, decisions, evidence, debt, and the next action.
The graph starts with the active goal, branches into a decision and implementation task, reconciles evidence and debt, then selects the next runnable task.
Every run starts from the current goal, known blockers, recent evidence, and a recommended next step. Completed work updates the shared state for the next agent or teammate.
Agents start with the current goal, not yesterday’s prompt.
Decisions, blockers, and evidence stay attached to the work.
Each run reports what changed, what was tested, and what should happen next.
Independent workstreams can run at the same time without stepping on the same node.
Works alongside your project tracker
Linear and Jira are useful for planning. Continuity sits closer to execution: it captures run output, evidence, blockers, and the next safe action for AI coding agents.
Status depends on humans remembering to update it.
Work state updates from agent runs, reported changes, evidence, blockers, and decisions.
Reasoning is scattered across comments, chats, transcripts, and local context.
Goals, decisions, tasks, blockers, follow-ups, and proof stay connected to the work.
A human still has to brief the agent and decide what is safe to do next.
Continuity recommends the next runnable work item and flags blocked or conflicting work.
Recent product proof
In a recent end-to-end run, Continuity turned a mobile CEO inbox request into graph placement, session specs, and parallel work kickoff — while preventing duplicate agents from taking the same graph node.
Three independent work runs were created and kicked off from the same intake flow.
Continuity converted a high-level request into structured work sessions agents could execute.
Parallel execution completed faster than running the same work sequentially in the smoke measurement.
A duplicate active work claim on the same graph node was rejected instead of allowing two agents to race.
Duplicate graph-node work was rejected with work_run_conflict.
Early founder access
For founders and small teams already using AI coding agents on production work. Continuity gives you shared context, clean handoffs, conflict prevention, and execution history across agent runs.
Designed to fit your workflow
No. Use your tracker for planning. Use Continuity to capture what agents actually changed, verified, blocked, and learned.
Prompts help one session. Continuity keeps state across many sessions, agents, and handoffs.
Yes. Independent lanes can run together while duplicate active work is blocked before it races.
Solo founders, technical CEOs, and small teams using coding agents for real product work.
Early access
If your team is already using coding agents and losing time to context recovery, duplicate work, stale handoffs, or unclear blockers, Continuity is built for you.