跳到正文
cccskills
免费在 GitHub 公开

progressive-disclosure

Refactor large DataRobot skill files by moving detailed content into directly linked reference files while preserving meaning. Use when a skill triggers context-window warnings, needs progressive disclosure, or should be chunked without changing guidance.

中文简介正在准备中。当前显示原文说明。

实测评分 未实测

查看安装方法

包含的文件(1)

  • SKILL.md5.1 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Progressive Disclosure

Use this repo-local meta skill to split large skills into a concise SKILL.md plus adjacent reference files. This guides content moves and reference creation; it is not a content rewrite pass.

Guardrails

  • Do not rewrite, reinterpret, simplify, or change the meaning of skill content.
  • Do not delete details because they seem verbose. Move details to reference files instead.
  • Do not change SDK guidance, commands, workflows, safety notes, or examples except to preserve links after moving content.
  • Do not edit plugin manifests, versions, CODEOWNERS, or unrelated repo metadata unless the user explicitly asks.
  • Do not create deeply nested references. Reference files should be directly linked from SKILL.md.
  • Do not move first-use trigger guidance, required safety constraints, or critical prerequisites out of SKILL.md.
  • If content needs substantive editing, stop and ask the user before making that change.

Target Shape

Keep SKILL.md focused on:

  • Frontmatter with accurate name and description
  • When to use the skill
  • Quick start or primary workflow
  • Key decision points and safety constraints
  • Direct links to reference files

Move long supporting content into files such as:

  • reference.md
  • examples.md
  • troubleshooting.md
  • platform-setup.md
  • sdk-operations.md
  • ci-cd-patterns.md

Choose names that match the moved section. Prefer a small number of meaningful files over many tiny fragments.

Use folders only when the supporting material has a clear type or grouping:

  • references/ for longer conceptual guides, examples, and workflow details
  • scripts/ for executable helpers that the skill may ask the agent to run
  • Language folders such as python/, typescript/, or r/ when a skill supports multiple programming languages

Keep these folders directly under the skill directory. SKILL.md should link directly to files inside them, such as references/workflow-details.md, without requiring the agent to follow a chain of references.

Workflow

  1. Inspect the target skill:

    • Read skills/<skill-name>/SKILL.md.
    • Check current line count and major headings.
    • Check current token warning/error status from task test:integration when available.
    • Identify sections that are detailed reference material, long examples, command recipes, troubleshooting tables, platform-specific variants, or API details.
  2. Propose the split before editing when the change is substantial:

    • List which sections will stay in SKILL.md.
    • List which sections will move and the destination reference file for each.
    • Keep a simple traceability map: original heading -> destination file -> replacement link text.
    • Confirm any ambiguous sections with the user.
  3. Move content with minimal transformation:

    • Preserve headings and body text as much as possible.
    • Adjust heading levels only when needed for valid Markdown structure.
    • Preserve code blocks exactly unless a relative link/path must change.
    • Keep warnings, prerequisites, and safety notes attached to the relevant content.
  4. Replace moved content with concise pointers:

    For detailed deployment configuration, see [deployment-reference.md](deployment-reference.md).
    

    Each pointer should explain when to read the reference, not summarize the whole reference.

  5. Keep links one level deep:

    • SKILL.md should link directly to every reference file the agent may need.
    • Reference files may link back to SKILL.md if useful.
    • Avoid chains like SKILL.md -> references/overview.md -> references/deep-reference.md.
  6. Validate:

    • Run task test:integration for structural checks.
    • Run task lint when tooling is available.
    • If the goal is reducing context-window warnings, rerun the relevant tests and compare warning output. This repo warns at an estimated 3300 tokens and errors at 6700 tokens.

Candidate Section Heuristics

Good move candidates:

  • Long examples that are not needed for initial task routing
  • Repeated setup variants by platform or provider
  • Detailed SDK method inventories
  • Troubleshooting catalogs
  • Extended CI/CD snippets
  • Long validation or debugging procedures

Usually keep in SKILL.md:

  • The first-pass decision tree
  • Required safety constraints
  • Minimal happy-path workflow
  • Short command checklist needed for most runs
  • Links to all supporting references

Diff Review Checklist

Before finishing, verify:

  • Every moved block still exists in a reference file.
  • SKILL.md links directly to each new reference file.
  • The frontmatter description still includes clear trigger scenarios.
  • The final response includes the traceability map for reviewer confidence.
  • No instruction, recommendation, or example changed.
  • No unrelated formatting churn was introduced.
  • Repo validation passes or any local tooling gap is clearly reported.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Use when the user wants to design, build, code, simulate, or deploy an AI agent (not a predictive model) to DataRobot; mentions agent_spec.yaml, dr-assist, datarobot-agent-assist, dress rehearsal, swarm simulation, or the DataRobot agent template; wants to scaffold a LangGraph, CrewAI, LlamaIndex, NAT, or Base agent targeting DataRobot; wants to add an MCP server, backend API, or React frontend to a DataRobot agent application; or uses the DataRobot CLI (dr) to build or deploy an agentic custom application; or wants to harden, stress-test, or battle-test an agent. Covers the full workflow: agent design, agent_spec.yaml authoring, dress-rehearsal simulation via the DataRobot LLM Gateway, adversarial swarm simulation, template-based coding, and deployment.

中文简介正在准备中。当前显示原文说明。

datarobot-oss/datarobot-agent-skills322026年10月23日 更新

Use when the user wants to design, build, code, or deploy an AI agent on DataRobot; mentions agent_spec.yaml, dress rehearsal, the DataRobot agent template, LangGraph, CrewAI, LlamaIndex, NAT, Base agents, MCP servers, backend APIs, custom frontends, or the DataRobot CLI.

中文简介正在准备中。当前显示原文说明。

datarobot-oss/datarobot-agent-skills322026年10月23日 更新

Use when an agent assist user wants to deploy their agent, put it on DataRobot, get a link to it, or share it with others, the agent can run as a single container, and the project has no `infra/Pulumi.yaml`. Not for agents built from the agent template, which deploy with Pulumi through datarobot-agent-assist. Users usually just say "deploy" or "deploy to DataRobot" and won't mention containers, workloads or the CLI.

中文简介正在准备中。当前显示原文说明。

datarobot-oss/datarobot-agent-skills322026年10月23日 更新

Use when the user wants to adversarially test, evaluate, or harden an implemented AI agent before deployment; mentions swarm simulation, attack testing, persistence testing, evaluation criteria, or eval_report.md.

中文简介正在准备中。当前显示原文说明。

datarobot-oss/datarobot-agent-skills322026年10月23日 更新

Use when the user wants to configure LLM integration for a DataRobot agent application. This skill helps to change LLM model, switch between the LLM integrations the project supports (LLM Gateway, a DataRobot-deployed LLM, an external provider, an LLM Blueprint, and any others its config declares), or set up provider credentials. The skill reads the project's .datarobot/cli/llm.yml for the real options, interviews the user, then runs sync_llm_env.py with the chosen values as CLI args to merge into .env.

中文简介正在准备中。当前显示原文说明。

datarobot-oss/datarobot-agent-skills322026年10月23日 更新

Guidance for setting up CI/CD pipelines for DataRobot application templates using GitLab, GitHub Actions, and Pulumi for infrastructure as code. Use when setting up CI/CD pipelines, configuring deployments, or managing infrastructure for DataRobot application templates.

中文简介正在准备中。当前显示原文说明。

datarobot-oss/datarobot-agent-skills322026年10月23日 更新

查看 datarobot-oss 的全部技能

报告此技能的问题