Kaplira FAQ — Architecture, Data, and Governance

Question 1

I already keep .md files for my agent. Why add Kaplira?

Accepted Answer

Kaplira works alongside your notes. It keeps accepted decisions and known errors linked to the files they affect, gives the agent the ones a governed task touches, checks planned edits against the task's scope when a pre-edit gate applies, and records what changed.

Question 2

Do I have to change how I work with Claude Code or Codex?

Accepted Answer

No. You keep your agent, editor, and terminal. The agent calls Kaplira's MCP tools to get the task's context and scope and to close the run with its evidence; you review the result in the Kaplira desktop app.

Question 3

Where do my code and architecture stay?

Accepted Answer

The active workspace, technical index, repository graph, and run records stay on your computer. A .kap file is a portable snapshot for export, migration, or backup. Kaplira does not hand custody of the repository to a cloud service.

Question 4

How does Kaplira keep agent changes within scope?

Accepted Answer

For work routed through Kaplira, a bounded contract and pre-edit gates check declared files against authorized scope. Review checks observed changes. Writes entirely outside the governed path cannot always be prevented in advance.

Question 5

Which coding agents work with Kaplira?

Accepted Answer

Kaplira works with MCP-capable executors including Claude Code, OpenAI Codex, and custom integrations. The executor implements while Kaplira selects context, grants scope, applies policy, and validates the result.

Question 6

What are safeguard and intervention receipts?

Accepted Answer

They are structured records of prevention decisions such as blocked scope expansion, missing required plans, or repeated attempts stopped by the anti-loop, including the policy, evidence, and result.

Architecture, data, and governance explained.

Question the page does not answer?