Source control
git diff, git log, git show, git status. Keeps every change but
trims unchanged context, omits binary blobs, and collapses long history.
What gets compressed
TokenShift recognizes the commands AI agents run most often and rewrites their output into a form the model can still reason about — but at a fraction of the token cost. The full original is always kept on the local device and can be fetched back at any time.
Categories of output we shrink
Section titled “Categories of output we shrink”Test output
pytest, jest, vitest, go test, cargo test. Keeps failing tests
with stack traces and assertion details; drops the wall of pass lines.
Build & lint
npm, cargo, go build, eslint, tsc. Keeps errors and warnings;
drops the noise around them.
Search
rg, grep, find. Groups results by file, trims duplicate matches.
Listings
ls, tree, git ls-files. Bounds depth, summarizes large directories.
Infrastructure
kubectl, docker, gh, aws, gcloud. Keeps the structure of
resources; drops boilerplate fields the model already knows.
What the agent sees
Section titled “What the agent sees”A compressed response always carries one extra line:
… tokenshift recover <hash>That hash points to the full output, stored on the local device. If the
compressed view ever isn’t enough, the agent runs tokenshift recover <hash>
and gets the original back — verbatim. The agent learns to do this on its own.
What we never compress
Section titled “What we never compress”- Anything the agent writes (your prompts, file edits, code) — TokenShift only touches command output, never input.
- Commands TokenShift doesn’t recognize — they pass through untouched.
- Local files — TokenShift never reads the filesystem outside of running commands you’ve asked the agent to run.