Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/visionTw/godot-ai-harnessWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/visiontw/godot-ai-harness/sync-learnings-to-harness)<a href="https://agentmods.dev/commands/visiontw/godot-ai-harness/sync-learnings-to-harness"><img src="https://agentmods.dev/badge/commands/visiontw/godot-ai-harness/sync-learnings-to-harness/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/visiontw/godot-ai-harness/sync-learnings-to-harness"><img src="https://agentmods.dev/badge/commands/visiontw/godot-ai-harness/sync-learnings-to-harness.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.01005 |
| Opus 5 | $0.00000 | $0.00502 |
| Sonnet 5 | $0.00000 | $0.00201 |
| Haiku 4.5 | $0.00000 | $0.00101 |
Grade A, and why
sync-learnings-to-harness scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 65 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/sync-learnings-to-harness
把业务仓最近学到的"可复用通用经验"反向沉淀到 godot-ai-harness，让其他业务仓也能受益。
触发时机（什么时候用这个命令）
- 每个 Phase 收尾时（建议作为 phase 关闭的最后一个动作）。
- 完成一次跨系统重构后。
- 解决了一个值得记录的非平凡踩坑后。
- 用户明确要求"这次的经验整理沉淀一下"时。
输入材料
- 最近 N 条 commit（默认 12 条；可配
/generate-harness-sync-brief先生成清单）。 - 最近会话中关键的 AI 问答与决策。
- 阶段开发日志
docs/phases/phase-*/development_log.md。
提炼标准（什么算"通用"）
| 通用（沉淀到 harness） | 项目专有（留在业务仓） |
|---|---|
Godot 引擎踩坑 → core/memory/godot_pitfalls.md |
玩法逻辑、卡牌/敌人数值、玩法专有 FSM |
harness/git/shell 踩坑 → core/memory/harness_ops_pitfalls.md |
项目专有的工程脚本路径 |
通用 GDScript 规范 → core/rules/godot-4-gdscript-standards.mdc |
项目专有命名/资源约束（如 project-roguecard.mdc） |
跨项目复用的 skill / command 模板 → core/skills/ / core/commands/ |
仅在某个项目玩法范畴下的步骤 |
与 Godot/MCP/Cursor 工作流相关的通用流程 → core/rules/godot-mcp-workflow.mdc 等 |
项目专有 smoke/regression 路径 |
判断口诀：换一个 Godot 4.x 项目还成立 → 通用；只有本项目世界观/玩法/资源结构成立 → 专有。
执行清单（推荐顺序）
- 生成基础清单：跑
./tools/harness/generate-harness-sync-brief.command（输出到docs/harness-sync-briefs/<date>.md）。 - 手工补"AI 问答复盘"段：把会话中关键决策、踩坑、通用方法补进去。
- 逐条分类：每个候选项标记"通用 / 专有"，在"可提炼到 Harness"和"留在项目仓库"两段下分别填位置。
- harness 改动：
cd vendor/godot-ai-harness- 按分类把内容写入对应位置（rules/skills/commands/memory/docs）
git add ... && git commit -m "..."git push origin main
- 业务仓推进 submodule 指针：
cd <business-repo>git add vendor/godot-ai-harnessgit commit -m "Bump godot-ai-harness to <hash>: <what>"git push
- 重新跑 bootstrap（可选，让本仓的
.cursor/_harness_*立刻拿到新内容）：./tools/harness/bootstrap.command
- 填同步执行记录：在
docs/harness-sync-briefs/<date>.md末尾补 harness commit hash 与 业务仓 commit hash。
提交顺序硬性约束
- 必须先 push harness，再 push 业务仓。
- 否则业务仓的 submodule 指针会指向远端不存在的 commit（CI 会失败，详见
harness_ops_pitfalls.mdH-002）。
输出归档
每次执行后：
docs/harness-sync-briefs/<date>.md：本次反向同步的完整清单与执行记录- harness 仓的 commit log：通用经验沉淀痕迹
- 业务仓的 commit log："Bump godot-ai-harness" 提交，便于回溯
与 /generate-harness-sync-brief 的关系
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 65 lines · 0 tokens per session scan A 7f6927305d73
sync-learnings-to-harness is a command published in the GitHub repository visionTw/godot-ai-harness (2 stars, last pushed 23d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,005 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
sim
Run the EmbodiedGen simulation rendering workflow with sim-cli.
cert-check
Review the current state against certification and platform compliance expectations.
combat-design
Design or revise combat rules, variables, readability, and tuning direction.
ui-flow-review
Review menus, HUD, navigation, and player flow from a UX perspective.
status
Show 3d-design team status and recent activity.
unity-optimize
Profile and optimize performance — uses MCP profiler for frame timing, memory, rendering stats. Identifies bottlenecks and applies fixes.