A shared engine supports many hosted, gateway, compatible, and local providers.
Agent Atlas / Coding agents
cline/cline
An open coding-agent engine presented through IDE extensions, a CLI, an SDK, a multi-agent task board, scheduled work, plugins, MCP, and checkpoints.
Coding agent
Full dossier
Project documented
Reviewed August 26, 2026
Intended operator
Developers supervising work in an IDE, terminal, automation, or task board.
Primary surfaces
VS Code and JetBrains / CLI and SDK / Kanban and scheduled agents
Tracked repository
Teaching mechanism
Plan and Act separate exploration from approved effects
The project documents a Plan mode for investigation and strategy, an Act mode for execution, per-action approvals, optional auto-approval, diffs, and reversible checkpoints.
Evidence limitProject documentation describes the control contract; TheGitReporter has not yet traced every surface through the shared engine.
Inspect the source
System Map
Eight inspection boundariesProject structure, rules, skills, session context, checkpoints, and persistent team state shape work.
File edits, terminal commands, browser actions, plugins, custom tools, and MCP execute across IDE and CLI hosts.
Per-action approval and optional auto-approval create distinct operator modes.
Diff review and checkpoints support modification and reversal; team and scheduled state persist across sessions.
Coordinator and specialist agents, worktrees, dependencies, and team state form the documented multi-agent surface.
Plan/Act, diffs, approvals, checkpoints, and task-board views expose different control moments.
Repository packages, evals, changelogs, docs, diffs, and checkpoints provide the inspection surface.
Learn Through This Project
10 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 5
Delegation Is a Distributed System
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
Latest Reporting
8 articles
Before Cline Replaces an Agent Hub, It Starts Saying No
August 23, 2026 — Cline’s Hub can now refuse new mutable work, wait for accepted sessions to settle, and replay a durable event stream to reconnecting clients before an upgrade swaps the service underneath them. cline/cline
Coding Agents Are Building Themselves—But Git Can Prove Only Part of ItAugust 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
Cline gives a finished agent answer a diff receiptAugust 9, 2026 — Cline’s restored “View Changes” control makes a completed agent turn point to a checkpointed file delta—not just its own verdict. cline/cline
An Approval Cannot Outlive the Conversation That Asked for ItAugust 2, 2026 — Cline fixed a small but revealing failure in its VS Code extension: change a message while a command approval is waiting, and the old run could keep waiting forever—then catch answers meant for the new one. cline/cline
The Harness Was Taking the Test TooJuly 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
An Agent’s Answer Begins in the HarnessJuly 26, 2026 — The model only sees the task after a runtime has edited its context, carried forward state, and decided what survived the last turn—so the first question after a bad run is what the harness made it see. QwenLM/qwen-code, cline/cline, block/goose, google-gemini/gemini-cli
Before You Blame the Model, Inspect What the Harness Made It SeeJuly 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
Cline: the supervised editor agent growing into an SDK and CLIJuly 26, 2026 — Cline's newest desktop work is not just another client: it moves chat toward a shared Hub that has to keep sessions, capabilities, and approvals legible across surfaces. cline/cline
Primary Sources
Open Questions
- Does the same approval and checkpoint contract survive IDE, CLI, headless, scheduled, and multi-agent modes?
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.