inlinr in your AI assistant
Ask your assistant how long the checkout took this week, log the call you just had without leaving the chat, or have it read a package's breaking changes before it upgrades. The inlinr command serves inlinr's answers to any MCP client, from your machine.
Get it on GitHubSet it up
Install the inlinr command. On Windows, run irm https://inlinr.com/install.ps1 | iex in PowerShell instead:
Terminalcurl -fsSL https://inlinr.com/install.sh | shConnect this machine, then open the link and approve with GitHub:
Terminalinlinr activateAdd the server to Claude Code:
Terminalclaude mcp add inlinr -- inlinr mcpOr to Cursor, Claude Desktop or another client, in its MCP settings:
Terminal{ "mcpServers": { "inlinr": { "command": "inlinr", "args": ["mcp"] } } }
What your assistant can ask
- Your time by project and language for a period, with assistant time and its cost
- A manual entry for a call or a meeting, logged from the chat
- The package.json it works on, checked for new releases, deprecations and advisories
- A package's releases and the breaking changes between two versions
- Your upgrades, the most pressing first, and the lines of your code that use a package
Good to know
- The package tools work without an account; your time and entries need inlinr activate.
- Finding where a package is used searches the files on your machine and sends nothing.
- Checking a package.json sends its four dependency blocks only, never its scripts or its name.
Never file contents, prompts or replies: the assistant sends inlinr only what a tool asks for. The full list of what is sent · Tracking Claude Code, Codex and Gemini CLI: terminal, desktop, IDE and cloud
Questions about your AI assistant
Which assistants can use it?
Does it need Pro?
Is it the same thing as the Claude Code plugin?
Read also
- IntegrationTime and token tracking for Claude CodeTrack Claude Code sessions per project: time, lines written, and the four token counts with their cost.Open
- ProductDependency trackingNew releases, deprecations and advisories for the packages your projects declare, with the time each upgrade took.Open
- AI usageReading a token bill: four prices, and the cache that is most of itFour counts, four prices, and a cache that re-reads the whole conversation every turn. How to read them without being off several times over.7 min read
Ask your assistant about your time
Free for solo use. Sign in with GitHub, add the server, and ask your first question.
No card to start · no source code leaves your machine