SKI
Voice coding for Claude Code and Codex — your agent talks back.
About SKI
SKI is voice coding for Claude Code, Codex and other agents — but it is not dictation. Your agent answers you out loud, like a real teammate, so you can build at the speed you think instead of the speed you type.
It sits ambiently on your desktop: hit a key, talk, and it works. You can even bring it into a meeting to build live or have it speak on your behalf. Runs locally on Mac and Windows, and it is free. A genuinely different way to drive an agent.
More Tools
Track your brand across ChatGPT, Claude, Gemini and Perplexity.
AI Search Console
AI Search Console is built for the new reality of GEO — generative engine optimization. It replaces manual AI visibility checks with repeatable data: brand mentions, rankings, share of voice, competitors, and cited sources across ChatGPT, Claude, Gemini, and Perplexity. You can analyze visibility at the individual prompt level, find content and citation gaps, and generate client-ready reports without spreadsheets or screenshots. If people are asking AI about your product, this is how you see what it says.
AI code review that catches product drift before it ships.
Prelint
Prelint reviews every pull request against your architecture decision records, docs, and past decisions — catching product drift before it merges rather than after it ships. When AI writes code at 10x speed, this is the check that makes sure it is the right code. On teams already running several AI reviewers, roughly 40% of the issues fixed before merge are caught by Prelint. If your agents ship fast but wander off-spec, it is a practical guardrail worth adding to the pipeline.
Agent-connected SEO research and audits from $10/month via MCP.
OpenSEO
OpenSEO gives your agent real SEO data instead of generic advice: keyword research, competitor research, backlinks, and site audits — starting at $10/month instead of the usual $100-plus subscriptions. Because it connects over MCP, you can work with your coding or writing agent to build an SEO strategy and produce content tailored to your company. A practical, low-cost data source for builders doing their own growth.
rtk
rtk is a CLI proxy that reduces LLM token consumption by 60–90% on common dev commands. It sits between your coding agent and the tools it runs, compressing noisy command output before it ever reaches the model — shipped as a single Rust binary with zero dependencies. Why it matters Agentic coding burns tokens fast, and a huge share of that is verbose tool output: build logs, test runs, file dumps. rtk trims that overhead so your agent spends its context window on signal, not noise — which means lower cost and longer, sharper sessions. Single Rust binary, zero dependencies 60–90% fewer tokens on common dev commands Drop-in proxy for agentic coding workflows Check it out on GitHub .
RubyLLM
RubyLLM is a delightful Ruby framework that gives you a single, elegant interface to every major AI provider — OpenAI, Anthropic, Gemini, and more. Chat, stream, embed, and call tools without juggling provider-specific SDKs.