# coding agents insights

Rangoon Insights reporting and guides tagged coding agents.

Canonical: https://rangoon.ai/insights/tags/coding-agents/

## Topics
- [Agent systems](https://rangoon.ai/insights/topics/agents/): Agent runtimes, handoffs, tools, and the boundaries that make them reviewable.
- [Governance](https://rangoon.ai/insights/topics/governance/): Policy, identity, authorization, approvals, and evidence for consequential work.
- [Infrastructure](https://rangoon.ai/insights/topics/infrastructure/): Model serving, sandboxes, tool services, and deployment boundaries.
- [Models](https://rangoon.ai/insights/topics/models/): Inference choices, local serving, compatibility, and operational tradeoffs.

## Stories
- [GLM-4.7 made preserved thinking a named control for multi-turn coding agents](https://rangoon.ai/insights/archive-zai-glm-4-7-december-2025/)
  Z.ai released GLM-4.7 on December 22, 2025, pairing open model weights with interleaved, preserved, and turn-level thinking controls for agent workflows.
- [Amazon introduced SWE-PolyBench to broaden coding-agent evaluation beyond Python repositories](https://rangoon.ai/insights/archive-amazon-swe-polybench-multilanguage-coding-benchmark-april-2025/)
  Amazon introduced SWE-PolyBench on April 23, 2025, with 2,110 repository issues across Java, JavaScript, TypeScript and Python.
- [OpenAI introduced SWE-Lancer to test coding agents against freelance software work](https://rangoon.ai/insights/archive-openai-swe-lancer-freelance-software-benchmark-february-2025/)
  OpenAI introduced SWE-Lancer on February 18, 2025, using freelance software tasks, end-to-end tests and manager-choice evaluations.
- [Devstral 2 paired code-agent weights with an execution-capable CLI](https://rangoon.ai/insights/archive-devstral-2-vibe-cli-december-2025/)
  Mistral released Devstral 2 on December 9, 2025, introducing 123B and 24B coding models with distinct licenses and the Mistral Vibe CLI.
- [Claude Sonnet 4.5 linked coding capability to checkpoints and agent operations](https://rangoon.ai/insights/archive-anthropic-claude-sonnet-4-5-september-2025/)
  Anthropic introduced Claude Sonnet 4.5 on September 29, 2025, alongside product changes including checkpoints, context editing, memory, and an Agent SDK.
- [SWE-Bench Pro brought private and long-horizon code tasks into agent evaluation](https://rangoon.ai/insights/archive-swe-bench-pro-september-2025/)
  Scale AI introduced SWE-Bench Pro on September 19, 2025, a 1,865-problem coding-agent benchmark spanning public, held-out, and commercial repositories.
- [Qwen3-Coder made repository-scale coding claims testable at the tool boundary](https://rangoon.ai/insights/archive-qwen3-coder-agentic-coding-july-2025/)
  Qwen released Qwen3-Coder on July 22, 2025, introducing a 480B-parameter MoE coding model and the Qwen Code command-line tool.
- [Cursor 1.0 turned code review and background execution into one editor release](https://rangoon.ai/insights/archive-cursor-1-0-june-2025/)
  Cursor’s June 4, 2025 1.0 release combined BugBot review, broader Background Agent access, project memories and one-click MCP setup.
- [Claude Code’s general availability joined a model launch to a developer workflow claim](https://rangoon.ai/insights/archive-claude-code-general-availability-may-2025/)
  Anthropic announced Claude Code general availability on May 22, 2025 alongside Claude 4, native IDE integrations and GitHub Actions tasks.
- [Google Jules public beta separated asynchronous coding from in-editor assistance](https://rangoon.ai/insights/archive-google-jules-public-beta-may-2025/)
  Google put Jules into public beta on May 20, 2025 as an asynchronous coding agent with cloud execution, plans and GitHub integration.
- [GitHub Copilot coding agent put a draft pull request at the center of agent delivery](https://rangoon.ai/insights/archive-github-copilot-coding-agent-may-2025/)
  GitHub introduced Copilot coding agent on May 19, 2025, with GitHub Actions execution, draft pull requests and existing repository policies.
- [AlphaEvolve made the evaluator the center of algorithm discovery](https://rangoon.ai/insights/archive-alphaevolve-evaluator-guided-algorithms/)
  Google DeepMind introduced AlphaEvolve in May 2025 as an evaluator-guided coding agent, making correctness checks and objective design central to claims of algorithmic improvement.


## More information

- [Documentation index](https://rangoon.ai/llms.txt)
- [AI access and policies](https://rangoon.ai/ai/)
- [Source repository](https://github.com/hypler-dev/rangoon)