eval-workflow
当用户要求为 agent 对话行为新增或修改 eval case、运行对话质量回归、评估 prompt、系统指令、skill name / description、工具集或模型升级的概率性质量影响，或补上普通测试无法覆盖的行为回归时触发。确定性代码正确性、普通单元/集成/E2E 测试与一般 TDD 实现不触发。
Skill by moesin-lab
Chalk Labs
Need help deploying eval-workflow?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/moesin-lab-eval-workflow
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: testing, productivity, agent-building
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/eval-workflow
What's inside
- Skill name
eval-workflow- Description
- 当用户要求为 agent 对话行为新增或修改 eval case、运行对话质量回归、评估 prompt、系统指令、skill name / description、工具集或模型升级的概率性质量影响，或补上普通测试无法覆盖的行为回归时触发。确定性代码正确性、普通单元/集成/E2E 测试与一般 TDD 实现不触发。
- License
- MIT
- Instructions
- 22 lines in SKILL.md · view on GitHub
Files (2)
- SKILL.md
- SKILL.md · harnesses/claude-code/
SKILL.md (first lines)
# eval-workflow（通用入口） 本 skill 是对话质量 eval 的薄入口，不定义 case、断言、记分或门槛。 ## 先读 完整读取： - [`docs/dev/testing/eval.md`](../../docs/dev/testing/eval.md) - [`docs/dev/testing/strategy.md`](../../docs/dev/testing/strategy.md) 涉及 fixture 时再读取 [`docs/dev/testing/fixtures.md`](../../docs/dev/testing/fixtures.md)。eval 资产与执行结果以上述 testing owner 为准。 ## 与普通 TDD 的边界 - 被测对象是 agent 的概率性行为质量时，进入本流程。 - 被测对象是确定性代码行为时，转交 `tdd-workflow`。 - 同一改动同时包含确定性实现和对话质量风险时，普通测试走 `tdd-workflow`，对话行为回归走本 skill；两类证据不互相替代。 当前 harness 有专用执行器时，使用 `harnesses/<harness>/SKILL.md` 中的能力映射。
From moesin-lab/agent-nexus, MIT.
More from moesin-lab/agent-nexus
- 0.521score
ci-recovery
moesin-lab
当 agent-nexus PR 的 GitHub Actions 或 GitHub 展示的 required check 已失败、取消或报红，用户要求查看日志、定位根因或修复 CI 门禁时触发。Checks 仍 pending、只需整体 readiness 判断，或本地测试失败但没有 PR check 时不触发。
SkillBFreshness: BClaude Code +9 - 0.521score
issue-intake
moesin-lab
当用户要求在 agent-nexus 仓库起草、创建或更新 GitHub Issue，把 bug、观察、想法、设计讨论或 review 发现转成 issue，或把 can-defer 项加入长期跟踪时触发。仅讨论、诊断或实现但未要求落 issue，以及关闭、重开、删除等 issue 生命周期操作不触发。
SkillBFreshness: BClaude Code +9 - 0.521score
development-workflow
moesin-lab
当用户要求在 agent-nexus 仓库实现功能、修复缺陷、修改代码或文档、调整配置或契约时触发。负责在动手前进入项目开发主路径，并按阶段加载需求澄清、文档判定、TDD、分支提交与 code review 的权威流程；不用于纯只读解释、状态查询或 code review。
SkillBFreshness: BClaude Code +9 - 0.521score
documentation-workflow
moesin-lab
当 agent-nexus 的改动需要判定是否新增或更新仓库文档，或用户要求起草、修改 ADR、spec、架构、流程、标准、测试或 docs/product 下的产品文档时触发。负责从 when-to-add-doc 进入 add-doc，并按判定结果路由到 ADR、spec 或其他文档…
SkillBFreshness: BClaude Code +9 - 0.521score
pr-feedback
moesin-lab
当用户要求检查、分类、处理或回复 agent-nexus PR 的 review comments、unresolved threads、requested changes、inline comments 或 reviewer 反馈时触发。独立 code review、创建或更新 PR、整体 readiness…
SkillBFreshness: BClaude Code +9 - 0.521score
pr-readiness
moesin-lab
当本仓库改动准备创建或更新 PR、声明 ready、交付、合并，或需要检查 PR 描述、review 证据、反馈闭环与整体门禁时触发；即使用户只说“提 PR”“可以交了”“看看能不能合”也应触发。不负责实现阶段 TDD，也不替代针对具体 unresolved review comments 或明确 CI…
SkillBFreshness: BClaude Code +9 - 0.521score
skill-setup
moesin-lab
当用户要求在 agent-nexus 仓库新增、修改、删除或重命名协作性 skill，调整 skills.manifest，或配置、修复仓库 skill 的 harness 挂接时触发。仅使用现有 skill、创建个人或全局 skill、修改普通项目文档时不触发。
SkillBFreshness: BClaude Code +9 - 0.521score
tdd-workflow
moesin-lab
当本仓库任务将新增功能、修复 bug、修改契约或改变可观察行为，并准备编写或修改实现与测试时触发；即使用户只说“实现”“修一下”或“补代码”也应触发。仅做测试流程路由，不用于纯探索、只读 review、无行为变化的文案修改或已经进入 PR readiness 的门禁检查。
SkillBFreshness: BClaude Code +9 - 0.521score
dependency-change
moesin-lab
当 agent-nexus 任务要求新增、升级、降级或移除外部依赖，修改或审计 workspace package 依赖关系、跨 package import / API 边界时触发；即使用户只说“装个库”“升级 SDK”“把这个包引进 daemon”或“检查跨包依赖是否违规”也应触发。不用于仅安装现有…
SkillBFreshness: BClaude Code +9 - 0.521score
release-workflow
moesin-lab
当用户要求 agent-nexus 发版、发布 npm 包、创建或推送发布 tag、验证 release candidate / CLI tarball，或修改发布 workflow 与 package 发布元数据后执行发布验证时触发。不用于普通 build/test、PR ready/merge、仅写…
SkillBFreshness: BClaude Code +9 - 0.521score
requirement-clarification
moesin-lab
当 agent-nexus 需求涉及多用户或公开面、授权访问控制、跨 package 契约、新增对外 API、修改既有 API 签名或 spec 字段、持久化、状态开关、配置 schema、改名或兼容性边界时，在 ADR、spec、测试或实现之前触发；用于发现用户原始表述之外的设计维度，不用于已明确方案之间的选型。
SkillBFreshness: BClaude Code +9 - 0.516score
subagent-usage
moesin-lab
当用户明确要求派发多个、若干或并行 subagent，或当前任务涉及跨文件探索、长日志/长输出分析、独立可验证子任务、独立 review / 第二意见、批量同质操作或研究类问题时触发。小型直接修改、需要持续用户澄清或高度共享上下文的任务不触发。本文件是 harness-neutral 薄入口；权威流程与标准在…
SkillBFreshness: BClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 14
- Last commit (this folder)
- Jul 31, 2026 (3 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/moesin-lab/agent-nexus
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from moesin-lab
- 0.521score
issue-intake
moesin-lab
当用户要求在 agent-nexus 仓库起草、创建或更新 GitHub Issue，把 bug、观察、想法、设计讨论或 review 发现转成 issue，或把 can-defer 项加入长期跟踪时触发。仅讨论、诊断或实现但未要求落 issue，以及关闭、重开、删除等 issue 生命周期操作不触发。
SkillBFreshness: BClaude Code +9 - 0.521score
development-workflow
moesin-lab
当用户要求在 agent-nexus 仓库实现功能、修复缺陷、修改代码或文档、调整配置或契约时触发。负责在动手前进入项目开发主路径，并按阶段加载需求澄清、文档判定、TDD、分支提交与 code review 的权威流程；不用于纯只读解释、状态查询或 code review。
SkillBFreshness: BClaude Code +9 - 0.521score
documentation-workflow
moesin-lab
当 agent-nexus 的改动需要判定是否新增或更新仓库文档，或用户要求起草、修改 ADR、spec、架构、流程、标准、测试或 docs/product 下的产品文档时触发。负责从 when-to-add-doc 进入 add-doc，并按判定结果路由到 ADR、spec 或其他文档…
SkillBFreshness: BClaude Code +9 - 0.521score
pr-feedback
moesin-lab
当用户要求检查、分类、处理或回复 agent-nexus PR 的 review comments、unresolved threads、requested changes、inline comments 或 reviewer 反馈时触发。独立 code review、创建或更新 PR、整体 readiness…
SkillBFreshness: BClaude Code +9 - 0.521score
pr-readiness
moesin-lab
当本仓库改动准备创建或更新 PR、声明 ready、交付、合并，或需要检查 PR 描述、review 证据、反馈闭环与整体门禁时触发；即使用户只说“提 PR”“可以交了”“看看能不能合”也应触发。不负责实现阶段 TDD，也不替代针对具体 unresolved review comments 或明确 CI…
SkillBFreshness: BClaude Code +9 - 0.521score
dependency-change
moesin-lab
当 agent-nexus 任务要求新增、升级、降级或移除外部依赖，修改或审计 workspace package 依赖关系、跨 package import / API 边界时触发；即使用户只说“装个库”“升级 SDK”“把这个包引进 daemon”或“检查跨包依赖是否违规”也应触发。不用于仅安装现有…
SkillBFreshness: BClaude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get eval-workflow running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
eval-workflow skill. Agent Mart. https://agentmart.dev/skills/moesin-lab-eval-workflow (last updated 2026-10-24)