Back to directory

ank

A command-line tool that stores a repository's tasks and architecture decisions in a project-local folder, so a coding agent can see what was decided and why, and claim, verify and hand back work.

ank

ank keeps tasks and architecture decisions attached to the code they constrain, in a folder inside the repository, and serves them through a single command-line tool any coding agent can call. Its stated aim is to give an agent access to context that lives in a tracker, a wiki or a decision thread, none of which it can otherwise read alongside the code.

The core workflow is a loop: read what binds the current work, claim a task and freeze its acceptance criterion, view the whole task, log what was learned, run its declared verifiers to record proof it was done, and release it back. It is explicitly not a project tracker or a wiki, and not a security boundary; it only guards against drift between decisions and the code.

It runs on Linux, macOS and Windows, needs git 2.34 or newer, and is licensed under Apache-2.0. Its storage format is not yet stable, though its command loop and exit codes are.

More in Agent memory

See all 450
  • A long technical guide to AI agent memory: why stateless LLMs need it, the different memory architectures from vector stores to knowledge graphs, and how Cognee implements one as an open-source memory layer.

  • A local, file-based long-term memory layer for a personal AI that never summarizes conversations, keeping every line with a timestamp and searching by time first and words second, rather than retrieving only similar-seeming chunks.

  • NEOTH

    deepwiki.com

    A local-first personal AI daemon in Rust with five memory tiers, consent-gated tools, a WASM plugin sandbox, multi-provider model routing and signed audit logs, aimed at people who want both a daily assistant and a serious coding helper.