rtk is a free, open-source Rust CLI proxy that intercepts shell commands run by coding agents and compresses their output before it enters the context window, claiming up to 90 percent fewer output bytes on supported commands.
rtk monetizes the least glamorous line in the agent budget, the command output nobody reads, and its own documentation admits the 90 percent headline is an output-bytes number, not a bill reduction. That admission is why I trust the project and discount the marketing at the same time.
What it is #
A single dependency-free Rust binary that sits between the agent and the shell: a pre-tool-use hook (or rules file) rewrites commands like git status to rtk git status, and rtk returns a filtered, grouped, truncated, and deduplicated version, with under 10 milliseconds of claimed overhead.
Coverage is per-command strategy: git status/diff/log, grep and ripgrep, find, test runners (pytest, cargo test, jest, vitest, playwright), linters and typecheckers, package managers, docker and kubectl, AWS CLI, and generic wrappers for anything else; git push becomes one line.
Full raw output is tee’d to disk when a filtered command fails, rtk proxy gives raw passthrough, and rtk gain reports measured savings.
Hooks ship for 16 tools including Claude Code, Codex, OpenCode, Cursor, Gemini CLI, Copilot, Windsurf, and Cline.
Apache-2.0, built by a small French team with a commercial Pro product alongside.
Status #
Young and hot: 82,266 stars, 5,223 forks, and 1,579 open issues and pull requests combined as of 2026-10-03, created 2026-01-22, pushed 2026-10-03. v0.50.0 went stable on 2026-09-24 and v0.51.0 followed on 2026-10-02, concluding that candidate train; a 0.51.1 candidate train is already running (rc.497 by 2026-10-03), and the project is pre-1.0. The site claims 18,000+ developers, a marketing figure, and the Show HN thread shows users posting their own measured savings.
Strengths #
- The mechanism is concrete and auditable: per-command filtering strategies in a binary you can read, not a prompt trick.
- Lossiness is mitigated by design: tee-on-failure preserves full output, and per-command opt-outs exist.
- Broad agent coverage with graceful degradation and a clean uninstall path.
- The docs measure and publish the limits of the savings claims, which is rarer than the savings claims themselves.
Cautions #
- Lossy output is the product: an agent reading a condensed diff or collapsed test list can misread state, and the changelog shows a steady stream of filter-correctness bugs.
- The hook only rewrites Bash tool calls, so built-in Read/Grep/Glob tools bypass filtering entirely, and unmatched commands pass through at zero savings; one user’s own log showed 39 percent on their busiest day.
- All savings numbers trace to the project or its users; no independent benchmark exists as of 2026-09-02.
- A crates.io name collision is documented in its own README, and the combined open issues and PRs total 1,579 as of 2026-10-03.
Pricing #
The CLI is free, Apache-2.0, no accounts. RTK Pro adds secret detection, org policies, observability, and sandboxed environments as a dedicated-tenant SaaS, self-hosted Helm chart, or on-prem deployment, with a 7-day trial; public per-seat prices are not published.
Compared to #
- Repomix: statically packs a whole repo for one prompt; rtk dynamically filters output during a live session, and the two are complementary.
- Plain CLI with no proxy: perfect fidelity and zero dependencies; choose it when exact output matters or your agent’s built-in tools dominate, rtk when long sessions are dominated by noisy commands on metered APIs.
- qmd: compresses which context enters the window from your documents; rtk compresses what the shell sends back, so they compose.
Bottom line #
Recommended for teams on metered APIs whose agent sessions are dominated by test, git, and search output, with rtk gain run for a week before believing any savings number.
Not for sessions driven by built-in Read/Grep tools, or anyone whose agents cannot tolerate condensed failure output.
Changes #
- 2026-08-30 - Created in the owner-directed star sweep as a Context engines note.
- 2026-09-05 - Corrected a factually wrong caution, the combined issue backlog is about 2,100, not star-sized.
- 2026-09-16 - Refreshed the volatile numbers (80,635 stars, 1,626 open issues and PRs, pushed 2026-09-15) and noted the v0.50.0 release-candidate train; latest stable remains v0.49.0.
- 2026-09-18 - Refreshed the volatile numbers (80,868 stars, 1,641 open issues and PRs, pushed 2026-09-18); stable remains v0.49.0, with the v0.50.0 release-candidate train still running (rc.442).
- 2026-09-20 - Refreshed the volatile numbers (81,065 stars, 5,138 forks, 1,572 open issues and PRs as of 2026-09-20, pushed 2026-09-19); stable remains v0.49.0 with the candidate train at rc.444, and the combined backlog dropped about 70 over two days.
- 2026-09-25 - Recorded v0.50.0 going stable (2026-09-24), concluding the release-candidate train, and refreshed the volatile numbers (81,660 stars, 5,166 forks, 1,576 open issues and PRs, pushed 2026-09-24).
- 2026-09-26 - Recorded the follow-up 0.51.0 release-candidate train starting (rc.465 as of 2026-09-26) and refreshed the volatile numbers (81,734 stars, 5,179 forks, 1,591 open issues and PRs, pushed 2026-09-26).
- 2026-09-27 - Refreshed the volatile numbers (81,785 stars, 5,185 forks, 1,543 open issues and PRs); stable remains v0.50.0 with the 0.51.0 candidate train at rc.468.
- 2026-10-03 - Recorded v0.51.0 going stable (2026-10-02), the candidate train concluded, and refreshed the volatile numbers (82,266 stars, 5,223 forks, 1,579 open issues and PRs, pushed 2026-10-03); a 0.51.1 candidate train is already at rc.497.
See also #
- Context Engines Feature Matrix - the category comparison this note joins
- Repomix - the static-packing complement
- Context Management Patterns - the manual practices rtk automates
- Model Selection for Coding Tasks - the per-token economics rtk attacks
References #
https://github.com/rtk-ai/rtk - repository, supported commands, agent integrations, license
https://raw.githubusercontent.com/rtk-ai/rtk/HEAD/README.md - the four filter strategies, 16 integrations, and the name-collision warning
https://raw.githubusercontent.com/rtk-ai/rtk/develop/docs/guide/resources/savings-explained.md - the project’s own methodology and the bytes-versus-bill caveat
https://github.com/rtk-ai/rtk/releases - release cadence and the filter-correctness bug history
https://www.rtk-ai.app - the marketing claims this note discounts
https://pro.rtk-ai.app - the Pro product surface and deployment models
https://news.ycombinator.com/item?id=47189599 - the thread with naming criticism and early user-posted gain stats