dsh-design-qa
让 DeepSeek Harness 里任何纯文本模型都能读图。识图是按需调用的 tool —— 图片不进主模型上下文,不看就不花钱;附 23 处缺陷的评测集,换模型可自测。
16 results
让 DeepSeek Harness 里任何纯文本模型都能读图。识图是按需调用的 tool —— 图片不进主模型上下文,不看就不花钱;附 23 处缺陷的评测集,换模型可自测。
OMK — Observe. Measure. Know. Evidence-backed knowledge changes for AI applications.
DSH 插件回归评测门禁:yaml 用例 + headless 驱动 + trace 断言 + baseline 门禁(eval_run / eval_gate)
Witness — the quality gate for agent-built web apps. Independent browser acceptance testing: CLI, MCP, AI-drafted checklists, visual regression, multi-browser. Agents self-test and pass; real browsers tell the truth.
Reproducible deterministic benchmark evidence for DSH tools and plugins
DSH web shell mobile adapter plugin: overlay sidebar/details drawers, responsive composer and settings below 768px, zero desktop regression. DSH Web 移动端适配插件:抽屉浮层化、输入栏与设置页响应式适配,桌面零回归。
DSH 插件风洞 — contract regression harness for DeepSeek Harness plugin bundles. Scripted-adapter driving, isolated child processes, session-event assertions.
Deterministic, CI-safe golden-output evaluation for DeepSeek Harness
Run one command N rounds and judge by median/distribution instead of a single run
Configurable user-confirmed lessons that help DeepSeek Harness avoid repeating mistakes.
dsh-reprolab — a complete bug-to-proof workflow for DeepSeek Harness: frozen repro tests, RED/GREEN/REGRESSION gates, and replayable evidence.
Deterministic regression tests for DSH policy decisions, evaluated without invoking the protected tool body.
Turn explicit coding-agent corrections into executable DeepSeek Harness regression tests.
Local-first visual acceptance and regression workbench for DeepSeek Harness
Evidence-first frontend release checks for DSH: real browser inspection, user flows, visual baselines, and reproducible reports.
DSH++ — a local-first control plane for DeepSeek Harness. Plugin security auditing and workflow regression testing for dsh, in a CLI plus a loopback web console.