DevAgentRadar → Repo Harness Rule Patch

Goal: update our repository's permanent harness instructions based on this release.

Update the file that tool actually reads. Do not dump everything into one file.

• Claude Code → CLAUDE.md plus .claude/ (skills, hooks, settings, agents, commands). It does not read AGENTS.md natively; start CLAUDE.md with @AGENTS.md.
• Codex, Copilot, Cursor, Factory Droid, Grok Build, Roo Code, Goose, OpenCode, Amp, Zed, Aider → AGENTS.md.
• Gemini CLI / Antigravity → GEMINI.md. AGENTS.md only if context.fileName is set.
• Cursor glob-scoped rules → .cursor/rules/*.mdc (plain .md is ignored), not a second constitution.
• Codex MCP → .codex/config.toml, not .mcp.json.
• Copilot extra instructions → .github/copilot-instructions.md; custom agents → .github/agents/.
• Windsurf → .windsurf/rules (do not assume AGENTS.md).
• Cline → .clinerules.
• Procedures → a skill (SKILL.md). Enforcement the model must not skip → a hook. Isolated roles → subagents.
• Do not fork the same rule into three tool files.

Context

Assistant: Zed AI Proposal: Enable low reasoning effort for DeepSeek V4 Flash and V4 Pro Summary: Zed v1.17.2 adds low reasoning effort for DeepSeek V4 Flash and V4 Pro, Gemini 3.7 Flash on Google AI, and an ask_user tool with selectable options or free-text. Default routine agent turns to low-effort DeepSeek and structured ask_user prompts. Primary source: https://github.com/zed-industries/zed/releases/tag/v1.17.2

Why it matters

v1.17.2 added low reasoning effort for DeepSeek V4 Flash and V4 Pro (#62577), Gemini 3.7 Flash for Google AI (#62670), and an ask_user tool so the Agent can present forms with selectable options, free-text, or both (#61497). OpenRouter reasoning continuity and request privacy also improved (#62652). Everyday coding-assistant turns often spend reasoning on clarification and easy edits; low effort on DeepSeek V4 plus ask_user forms keeps that work cheap and explicit. Use Gemini 3.7 Flash as a second low-latency Google AI option, and keep OpenRouter in the path when you need the continuity and privacy improvements.

Suggested actions

1. Open Zed Agent settings and set reasoning effort to low for DeepSeek V4 Flash and DeepSeek V4 Pro; use those models for routine edits, tests, and small refactors.
2. Add this to your Zed Agent system prompt: When you need input, call ask_user with selectable options, free-text, or both. Never ask a multiple-choice question only in the chat transcript.
3. In the Zed model picker, select Gemini 3.7 Flash under Google AI for fast turns that should not use DeepSeek.
4. Send multi-step DeepSeek agent threads through OpenRouter to pick up v1.17.2 reasoning continuity and request privacy.

Config surfaces this release may change

• context and compaction settings — check your repo before applying

After you finish

Do not report this as applied to DevAgentRadar. You cannot write the visitor's loop.

Tell the human: open https://devagentradar.com/proposals/zed-ai-v1-17-2-enable-low-reasoning-effort-for-deepseek-v4-flash-and-v4- and mark Applied, Skipped, or Failed. Proposal id: 7a8e1de2-c70d-4a13-9427-b4a0be8a5b70

Instructions:

1. Read the existing context files and settings for the tools in this repo.
2. Draft an explicit Git diff. Update ONLY agent harness rules or tool configuration.
3. Call out deprecated flags, obsolete habits, or changed permission boundaries.
4. Do not touch application logic.

Start with the proposed diff and say which layer and which tool's file it belongs in.