How It Works

Keshro turns projects into executable work.

Plan, execute, and track multi-step engineering work with agent coordination, durable execution state, and rollback built in.

The in-agent loop

/keshro or keshro create
review the generated plan
run keshro continue
watch status, notes, rollback

Step 1

Start from your agent or the CLI

Invoke Keshro from Claude Code, Codex, another supported agent, or the CLI. Prefer the agent flow for normal use, but every step is also available directly in the CLI.

Step 2

Keshro turns the request into execution state

Keshro creates real execution state: tasks, dependencies, acceptance criteria, risks, blockers, and issue links.

Step 3

You review first, then the agent executes

After review, run `keshro continue` or `/keshro`. Keshro picks the next work and fans out safe parallel tasks into isolated worktrees.

Step 4

Stay in control while work is running

Inspect status, see blockers, audit decisions, and roll back to checkpoints when needed.

Under the hood

What Keshro is doing while your agent is coding

Keshro handles the coordination layer so the agent can stay on implementation.

Topical context routing

Task 3 learns about your IAM config. Task 12 also touches IAM. Task 12 automatically gets task 3's discoveries, even if tasks 4-11 were unrelated.

Isolated worktrees

Each parallel agent gets its own git worktree. No merge conflicts, no stepping on each other's changes. Work merges cleanly when tasks complete.

Git checkpoints

Auto-commit before every task. If an agent goes sideways, roll back to exactly the state before it started. One command.

Validation gates

Agents must pass linting, tests, and syntax checks before they can mark a task done. No silent failures that pile up downstream.

Task splitting

When the scheduler hits a ready parallelizable task, it auto-splits it into child tasks. One agent starts the first child and other agents pick up the siblings in parallel.

Cross-project learning

Past project patterns are embedded and indexed. New projects with similar stacks get informed by what worked before and what didn't.

During execution

Keshro keeps the plan live

As work starts, blocks, or finishes, Keshro keeps task state, notes, and artifacts current.

Keshro carries context forward

Discoveries from finished tasks route into later related tasks instead of getting lost in chat history.

Keshro constrains risk

Checkpoints, isolated worktrees, and validation gates make execution reversible and controlled.

The fastest way to judge Keshro is to try a real project

Bring a project you're already working on. If the plan is useful, the execution will be too.