DiagForge Bootstrap

DiagForge Bootstrap

This is a lightweight onboarding 技能 for the DiagForge 仓库.

It is not the full DiagForge 系统. Its job is to 图形界面de an 代理 to the correct GitHub 仓库, documents, smoke test, and execution flow.

DiagForge itself is an 代理-driven closed loop built on top of Microsoft Visio. Its goal is to turn reference figures into directly editable diagram as设置s by helping 代理s operate Visio more like a capable human user rather than as a blind API caller.

What this 技能 can do

This 技能 can help an 代理:

understand what DiagForge is trying to achieve find the correct GitHub 仓库 and entry documents avoid random first-运行 behavior and jump into the intended 工作流 运行 the canonical cold-启动 smoke test 启动 reproducing reference figures through the DiagForge Visio loop move toward a 结果 that a human can continue editing directly in Visio Typical outcomes

After using this 技能, an 代理 should be able to:

explAIn the DiagForge 工作流 clearly bootstrap itself into the 仓库 with the correct read order 验证 that the Visio bridge and execution path are working begin work on figure reproduction with better layer awareness help produce editable .vsdx 输出s instead of dead image copies What this 技能 is for

Use this 技能 when an 代理 needs to:

find the DiagForge source 仓库 understand the top-level architecture quickly avoid free-form blind retries 运行 the canonical cold-启动 smoke test begin work in the correct layer When to use it

Use this 技能 when:

an 代理 is entering DiagForge for the first time a new 环境 needs to be 验证d before real drawing work a user wants an 代理 to help reproduce a figure through Visio the goal is not only to look similar to the reference, but to obtAIn a directly editable diagram as设置 What this 技能 is not

This 技能 does not bundle the whole 仓库. It does not include Visio bridge code, benchmark PNGs, or 运行time artifacts.

The full project lives in the GitHub 仓库:

https://github.com/qweadzchn/DiagForge

Recommended 工作流 Clone the GitHub 仓库 locally. Read the cold-启动 entry documents. 运行 the canonical smoke test before doing open-ended drawing work. Only then move on to real jobs or 系统 improvements. Clone the 仓库 git clone git@github.com:qweadzchn/DiagForge.git cd DiagForge

If SSH is not avAIlable, use HTTPS instead.

Read order

Read these files first:

代理_启动_HERE.md 代理_图形界面DE.md 获取_启动ED.md docs/human/设置up/代理_COLD_启动_SMOKE_TEST.md MODE_POLICY.md Canonical smoke test

From the repo root:

python 设置up\prepare_smoke_test.py --config 设置up\examples\smoke-test-输入png-1.json python 设置up\运行_draw_job.py --config 设置up\examples\smoke-test-输入png-1.json python 设置up\执行_drawdsl.py --config 设置up\examples\smoke-test-输入png-1.json --round 1 --save-final

Expected 输出s:

输出Preview/smoke-输入png-1/round-01.png 输出Editable/1_smoke_test_final.vsdx Routing rule

When working inside DiagForge:

if the issue is round-specific, keep it in review artifacts if it looks structural but still needs 验证, write a proposal if it is already reusable experience, promote it into a lesson if the 分享d fix is clear, 补丁 the owning layer directly Where to go next

See:

README.md CONTRIBUTING.md docs/architecture/FEEDBACK_PROMOTION_LOOP.md