swob

Adoption verdict: swob is worth a cautious trial for developers managing several AI coding tools, but the evidence does not justify a strong quality claim. The sample contains one GitHub topic entry and two X posts: the repository shows 24 stars and 2 forks, while the X posts total about 2,224 views. That demonstrates attention, not reliability, compatibility, or proven productivity gains. Discussion quality is limited: there are no hands-on tests, tutorials, long-form comparisons, or substantial user feedback.

The GitHub description presents swob as a local-first Git graph and session debugger for Claude Code, Codex, Cursor, OpenCode, and other AI coding harnesses. Its apparent job is to let developers inspect, manage, and analyze session history and lineage, connecting the AI coding process with Git changes. It is better understood as a session-record and debugging layer than as another coding agent. An appinn X post additionally claims cross-harness and cross-device aggregation, smart organization, insight, token statistics, and efficiency measurement; those are social-media claims, not independently tested guarantees.