Code search for coding agents: ranked, exact excerpts with line numbers.
io.github.bartlomein/oko β Model Context Protocol (MCP) Server
This MCP server provides code search for coding agents. It returns ranked results and supports exact excerpts, including line numbers, to help identify relevant code segments from a repository.
π οΈ Key Features
Code search for coding agents
Ranked search results
Exact excerpts
Excerpts include line numbers
π Use Cases
Finding relevant code locations for automated coding agents
Retrieving exact code excerpts with line-number context for review or reuse
β‘ Developer Benefits
Easier navigation from search results to exact code lines
More precise matches via exact excerpt retrieval
β οΈ Limitations
No additional capabilities (e.g., specific tools, integrations, or configuration) are described in the provided data.
Help your coding agent find code faster and use fewer tokens.
Oko gives Claude Code, Codex, Cursor, OpenCode and Pi a better code search. When
you ask your agent about your code, Oko has already found the relevant source,
so the agent spends fewer rounds searching and reading files it does not need.
Nine coding tasks on Astro, HTTPX and ripgrep, three runs each, with and without Oko. Results Β· Method
Faster: about half the time in Claude Code, OpenCode and Pi; a quarter less in Codex.
Cheaper: 33β44% fewer tokens with every tool we measured.
Fewer detours: 65β89% fewer tool calls.
Better ranking: puts the right file first more often than any published method on Agent Retrieval Bench.
Local first: search runs on your machine; only the question and selected snippets go to the ranker (privacy).
Quick start
sh
brew install bartlomein/tap/oko
cd your-project
oko setup --client claude # or codex, cursor, opencode, pi, all
Without Homebrew: curl -fsSL https://raw.githubusercontent.com/bartlomein/oko/main/install.sh | sh.
Then use your coding tool as usual. Setup asks for a TypeSafe AI
key the first time, connects your tool to Oko for this project, and checks the
connection. Run it once per project; oko upgrade keeps everything current
(with Homebrew, run brew upgrade oko first).
Codex: approve Oko's hook once
The first time you open the project, Codex shows Hooks need review. Choose
Trust all and continue. If you skipped it, /hooks brings it back.
Cursor: turn Oko on once
Cursor adds a project's tools switched off. Open Customize > MCPs, click
oko and turn it on.
Pi: trust the project
Pi 1.0 or later. Trust the project when Pi asks, then check /mcp.
Pi setup.
Claude Code, OpenCode: nothing more to do
Start a new session after setup.
Platforms, manual install, no-key mode
macOS (Apple Silicon and Intel) and Linux (ARM64 and x64, glibc 2.35+). The
installer puts oko in ~/.local/bin; add it to your PATH if needed.
Other install options Β· Connect a tool by hand Β·
Keys and configuration. For search without a key,
add --no-jev: local keyword ranking only.
Results
The same task given to the same coding tool twice, once with its own search and
once with Oko, in the same run. Means per session.
Coding tool
Time
Tokens
Tool calls
Tasks passed (without β with)
Claude Code
47% less 12.1 β 6.4 s
44% less 24,271 β 13,627
89% less 3.5 β 0.4
19 β 20 of 20
Codex
26% less 13.8 β 10.2 s
43% less 45,395 β 26,004
80% less 3.0 β 0.6
20 β 20 of 20
OpenCode
47% less 18.0 β 9.6 s
43% less 13,196 β 7,528
87% less 3.7 β 0.5
21 β 21 of 21
Pi
51% less 16.7 β 8.2 s
33% less 9,737 β 6,495
85% less 3.6 β 0.5
20 β 21 of 21
Cursor
6% less 52.3 β 49.2 s
40% less 182,940 β 109,520
65% less 8.0 β 2.8
26 β 24 of 27
Models: Claude Sonnet 5.5 (Claude Code), GPT-6.1 Sol (Codex, OpenCode, Pi), Grok 4.7 High (Cursor), at their default or low effort. Cursor's row is from its own run on the previous release; the others are from October 2026. Each tool counts tokens its own way, so compare within a row. Small benchmark; your results will vary. Method, per-task results and caveats Β· raw data.
848 real issues: rank the code fixing agents needed
One search ranks it within the range of agents that explore for many turns: 0.82
Agents: 0.72β0.95
Where Oko helps less
Cursor's own search is already fast, so Oko saves less time there: mostly tokens and tool calls.
Duplicated code can mislead it. When two files hold identical copies of a function, Oko may show the wrong one first; one Astro edit task in our benchmark fails this way.
Big changes across many files: one search returns five places, so an issue touching many files still needs the agent's own exploring.
Use it from your terminal
sh
oko ask "where is authentication handled?"
oko ask "how are retries handled?" --json
oko stats # what ranking has cost you
Oko searches your project locally (keywords, symbols and file structure), then
asks Jev, TypeSafe AI's relevance model, which of the
candidates actually answer the question, and returns those excerpts with exact
line numbers. oko setup connects it to your tool as an MCP server, adds short
guidance to the instructions your tool reads, and installs a hook that answers a
code question as you send it.
What setup changes Β· Privacy.