Agent Atlas / Coding agents

aaif-goose/goose

A local, general-purpose agent with desktop, CLI, and API surfaces, broad provider support, and an extension model built around MCP.

Coding agent Full dossier Project documented Reviewed August 26, 2026

Intended operator

People using a local agent for coding, research, automation, and other tool-using work.

Primary surfaces

desktop app / CLI / embeddable API

Tracked repository

aaif-goose/goose

Teaching mechanism

MCP extensions separate the local agent from its external capabilities

The project documents a local Rust agent that connects provider choices and external capabilities through standard extension contracts instead of baking every integration into one model path.

Evidence limitREADME-level extension counts do not prove consistent credential scope, approval behavior, or reliability across extensions.

Inspect the source

System Map

Eight inspection boundaries

Model boundary

Many hosted and local providers can sit behind the same local agent surfaces.

Context and memory

Local sessions and extension-provided context shape the run; retention details require deeper inspection.

Tools and execution

MCP extensions supply tools for coding and broader workflows to a Rust runtime.

Permissions and isolation

Local execution makes host authority and each extension's credentials central inspection questions.

Sessions and recovery

Desktop, CLI, and API continuity require a shared session contract that remains to be traced.

Delegation

The current profile does not claim a worker hierarchy without deeper source inspection.

Human control

Desktop and CLI surfaces give the operator different ways to inspect and steer local work.

Evidence surface

The public Rust source, self-test configuration, diagnostics, releases, and docs provide the starting receipts.

Learn Through This Project

9 lessons

Chapter 1 Before the First Model Request Chapter 2 The Invisible Prompt Chapter 3 Tool Calls: Contract, Authorization, Execution Chapter 4 Loops, Workflows, and Recovery Chapter 6 Memory Is a Typed Storage Contract Chapter 7 The Approval Trap Chapter 8 The Human Control Loop Chapter 9 Trust Must Make Its Case Chapter 10 The Future Has Three Branches

Compare Its Boundaries

7 labs

Comparison labTool Contracts Comparison labGoverned Memory Comparison labApproval Boundaries Comparison labHuman Workbench Comparison labSessions And Recovery Comparison labEvidence And Evaluation Comparison labMaturity Signals

Latest Reporting

7 articles

A Child Agent’s Approval Request Shouldn’t Hide in Another Tab

September 2, 2026 — Openwork now brings a child session’s blocked permission into the parent task; fresh Goose state-machine work shows why seeing the ask is only half the control. different-ai/openwork, aaif-goose/goose

Coding Agents Are Building Themselves—But Git Can Prove Only Part of It

August 11, 2026 — Across 40,345 unique commits in ten public agent codebases, coding-agent credit becomes sharply more visible in 2026—and the missing receipts matter as much as the names that appear. openai/codex, google-gemini/gemini-cli, anomalyco/opencode, QwenLM/qwen-code, cline/cline, aaif-goose/goose, OpenHands/OpenHands, charmbracelet/crush, mistralai/mistral-vibe, Kilo-Org/kilocode

API Discovery: Let an ACP Session Reply Before You Light Up the Client

August 10, 2026 — The Agent Client Protocol’s public TypeScript SDK gives a client a newSession() promise; a fresh Goose change shows why that reply is a useful readiness boundary. aaif-goose/goose

The Harness Was Taking the Test Too

July 30, 2026 — OpenAI says two context choices lifted GPT-5.6 Sol from 13.3% to 38.3% on the ARC-AGI-3 public set. Codex source shows those “settings” are really a production memory system—and rival harnesses reveal how much of it can be copied. openai/codex, pydantic/pydantic-ai, earendil-works/pi, NousResearch/hermes-agent, mistralai/mistral-vibe, anomalyco/opencode, QwenLM/qwen-code, google-gemini/gemini-cli, cline/cline, aaif-goose/goose, OpenHands/OpenHands

Before Goose Runs a New Model, Pick Its Permission Mode

July 27, 2026 — Goose has mapped Claude Opus 5 into its runtime. Before a new model reaches local tools, the operator still has to decide whether the session runs automatically, asks first, uses risk-based approval, or stays in chat. aaif-goose/goose

Goose: one open agent runtime across models, tools, and local execution

July 27, 2026 — A fresh Goose provider change adds Claude Opus 5 and adaptive-thinking support; the more consequential choice is still whether that model may act through local tools without asking first. aaif-goose/goose

Before You Blame the Model, Inspect What the Harness Made It See

July 26, 2026 — A model's next move is shaped by the tool output, trust state, history, and recovery rules a harness hands it—so a bad run is not automatically a bad model. QwenLM/qwen-code, cline/cline, aaif-goose/goose, google-gemini/gemini-cli

Primary Sources

goose repository

Primary public local-agent and extension source.

Open Questions

Can every extension make its credentials, approval requirements, and side effects legible across desktop, CLI, and API use?

Evidence state describes what this Atlas profile can support today. It is not a product score and it is not evidence that uninspected capabilities are absent.