Choose theme
AI Resources
goose
goose is an AI agent available as a desktop app, CLI, and API. It uses a configured model and connected tools to work on code, files, research, and other workflows.
The project belongs to the Agentic AI Foundation under the Linux Foundation. Recipes package a task's prompts, extensions, and settings for reuse. Use this as a first read, not a recommendation. Open the original project before trusting details like terms, limits, privacy, cost, setup, or safety.
What it is
An agent that uses tools
goose connects model responses to local or external tools, including MCP extensions.
Why it stands out
Reusable workflow recipes
Recipes keep instructions and settings together, with parameters for inputs that change between runs.
Availability
Desktop, CLI, and API
Official documentation covers installation, provider configuration, extensions, and recipe use.
Why it matters
What makes it useful
For a recurring code-review task, a Recipe can keep the review instructions and tool setup while accepting a new language or focus as parameters. The documentation includes that example, showing how to reuse the setup without rewriting the task each time.
What to know
Where it fits
goose sits between your chosen model and the tools doing the work. Extensions determine what it can reach; Recipes describe the workflow. Experimental subrecipes run parts of a task in separate sessions with their own context and return their outputs to the main recipe.
Notable points
What stands out
A Recipe can define a JSON response schema for its final output, giving a downstream script named fields to parse. A file parameter inserts the file's contents into the prompt, not just its path. Matching the output schema does not establish that the generated answer is factually correct.
Before using
What to review
Configure a model provider and inspect the extensions the workflow needs.
Connected hosted models handle the material sent to them under their own terms; running goose locally does not keep those requests on the computer.
Subrecipes are experimental. Their separate conversation context is not a computer or file-permission boundary.
Session exports retain work outside the app: JSON keeps session data, while Markdown provides a readable conversation. Token and cost displays can help track usage; public-price cost estimates are not the provider bill.
The project documents risks from running code and following instructions in untrusted content; connected tools can act on your computer.
The official goose review advisory affects CLI versions below 1.44.0 and names 1.44.0 as the fix. Check the installed version if using that command.
The later 1.53.0 release includes another review-command fix and remote MCP redirect checks. Check the installed version and current security notes before reviewing unfamiliar repositories or connecting remote MCP services; the older 1.44.0 fix does not cover these later changes.
Reader fit
Who may find it relevant
People who repeat tasks that need both model responses and tool actions.
Developers connecting agent output to scripts through structured recipe responses.
Users prepared to manage model settings and tool access.
Editorial note
Why LifeHubber lists it
A recurring job may need separate review, analysis and writing steps, rather than one growing conversation. goose can keep those steps in subrecipes that receive inputs, work in their own conversation context and return a result to the main workflow. It is worth exploring when you want to reuse that division of work, with each step's tools and model settings described in its recipe.
Source links
Source materials
Reader note
Before relying on this entry
LifeHubber lists entries to help readers inspect AI projects, not to endorse them or prove they are safe, suitable, accurate, maintained, or right for a specific use. We do not verify every entry in depth. Before relying on anything listed, review the original materials, terms, privacy practices, limits, and risks that matter for your situation.
What to explore next
Set the boundaries around a tool-using workflow.
A Recipe describes the task, but its connected tools determine what goose can reach. Continue with practical access, approval, stopping, and recovery checks before a workflow acts on important files or accounts.
More in AI Agents
Keep browsing this category
Explore more AI agent projects.
Paperclip
paperclipai/paperclip
A self-hosted server and dashboard for coordinating agent teams through companies, goals, roles, issues, heartbeats, budgets, approvals, and persistent activity records.
Papermorph
DozenTwelve/Papermorph
A coding-agent skill and web player for turning textbook PDFs into narrated interactive lessons and stories into paper-puppet films, with static-book output and optional video or LMS delivery.
Vane
ItzCrazyKns/Vane
A self-hostable AI answering engine for configurable search-style workflows, with local and cloud model providers, SearxNG-backed web search, cited sources, file uploads, and Docker setup paths.
For project maintainers
Listed here? You can use the badge.
If you maintain a project with a current LifeHubber listing, you may add the optional “Listed on LifeHubber AI Resources” badge to its README, docs, or website. No introduction or permission request is needed.
