docs(ai): converge external wording to AI-assisted + human review, not full-auto

Per the architecture-review redline (EIP3 决策会纪要: AI claims limited to 'AI-assisted, human-reviewed'; never promise 'AI full-auto development'), soften endpoint @Operation summaries, the default task objective, and AI docs that read as committed full-automation. Endpoint paths/method names unchanged (API contract). Added a capability-boundary disclaimer at the top of AI_AGENT_AUTO_DEVELOPMENT.md. 33 platform-ai tests green.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-06-19 16:39:45 +08:00
parent c379f5b1ea
commit 9aea6c1cd0
5 changed files with 10 additions and 8 deletions

View File

@@ -1,6 +1,8 @@
# AI Agent 全自动开发运行说明
# AI Agent 辅助开发运行说明(生成草稿 + 本地验证 + 人工复核发布)
`platform-ai` 的自动开发入口是 `POST /api/ai/features/auto-develop`。它只在本地环境执行草稿生成、本地落库、smoke 和自动修复;线上变更只输出人工部署包,不连接线上环境执行。
> **能力边界(对外口径,红线)**:本能力是「**AI 辅助生成草稿 + 人工复核发布**」,**不是** AI 全自动开发。当前只覆盖最简单的 CRUD recipe,存量复杂场景未取得数据;按《EIP3架构决策会纪要》红线,**严禁在报价/合同中承诺「AI 全自动开发 / 人力大幅下降」**。线上变更一律输出人工部署包、由人工复核后发布。
`platform-ai` 的辅助开发入口是 `POST /api/ai/features/auto-develop`。它只在本地环境执行草稿生成、本地落库、smoke 和自动修复;线上变更只输出人工部署包,不连接线上环境执行,须人工复核后发布。
## 本地运行配置