Agent SkillsHouseofmvps/ultraship › staying-current

staying-current

GitHub

规范Agent在回答版本敏感问题(如API、价格、模型ID)时,必须通过context7或Web搜索验证最新信息,禁止仅凭训练数据记忆作答,确保准确性。

skills/staying-current/SKILL.md Houseofmvps/ultraship

Trigger Scenarios

涉及库/框架/SDK API签名或配置 询问软件包版本、兼容性或弃用状态 查询CLI参数、模型ID或定价 包含'最新'、'当前'等时间敏感词汇

Install

npx skills add Houseofmvps/ultraship --skill staying-current -g -y
More Options

Use without installing

npx skills use Houseofmvps/ultraship@staying-current

指定 Agent (Claude Code)

npx skills add Houseofmvps/ultraship --skill staying-current -a claude-code -g -y

安装 repo 全部 skill

npx skills add Houseofmvps/ultraship --all -g -y

预览 repo 内 skill

npx skills add Houseofmvps/ultraship --list

SKILL.md

Frontmatter
{
    "name": "staying-current",
    "description": "Use whenever answering anything version-sensitive — library\/framework\/SDK APIs, package versions, model IDs, pricing, CLI flags, config, or \"latest\/newest\" anything. Verify against current sources instead of training data.",
    "allowed-tools": "Bash, Read, WebSearch, WebFetch"
}

Staying Current

Training data has a cutoff. Libraries ship breaking changes, prices change, model IDs get renamed, and APIs get deprecated after that cutoff. Answering version-sensitive questions from memory is the single most common way an otherwise-correct agent ships wrong code. This skill is the standing rule for not doing that.

The Ultraship Currency Guard hook (UserPromptSubmit) already fires on every prompt and injects a reminder when it detects version-sensitive language. This skill is what you do when that fires — or any time you're about to state a fact whose correct answer changes over time.

The rule

Before stating any of the following, verify it against a current source. Never answer from training data alone:

  • Library / framework / SDK API signatures, imports, hooks, options
  • Package versions, compatibility, and what's deprecated
  • CLI flags and config file schemas
  • Model IDs, model names, context windows, and capabilities
  • Pricing, rate limits, free-tier limits, quotas
  • "Latest", "newest", "current", "recommended" anything
  • Release notes, changelogs, migration steps

Where to verify

Source Use for
context7 MCP (resolve-library-idquery-docs) Library/framework/SDK documentation and API syntax. This is the primary source for code. Use it even when you think you know the answer.
WebSearch / WebFetch Versions, pricing, model IDs, release notes, deprecations, anything not in a library's docs. The current month is the search context — say "2026" in queries when recency matters.
The project's lockfile (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, Cargo.lock) The exact version actually installed here — always check this before assuming a version.

How to apply

  1. Detect. If the question touches anything in the rule above, do not answer yet.
  2. Check what's installed. Read the lockfile/manifest to learn the real version in this project.
  3. Pull current docs. Resolve the library on context7 and query the specific API. For non-library facts (pricing, model IDs), WebSearch then WebFetch the authoritative page.
  4. Answer from the source, and cite it. Quote the version/date you verified against.
  5. If you can't verify, say so. "I couldn't confirm this against a current source" is correct. A confident wrong API is not.

Anti-patterns

  • ❌ Writing an import or hook call from memory for a library you haven't checked this session.
  • ❌ Quoting a price or model ID without fetching the current page.
  • ❌ Assuming the latest major version — projects pin old ones; read the lockfile.
  • ❌ Treating a recalled Ultraship memory as current — memories are point-in-time; re-verify file paths, flags, and versions before acting on them.

Note on Claude/Anthropic specifically

When the task involves Claude, the Anthropic API, model IDs, or pricing, the same rule applies with extra force — these change frequently. Verify model IDs and pricing against current Anthropic docs before quoting them.

Version History

  • ed232cb Current 2026-07-24 16:16

Same Skill Collection

skills/a11y/SKILL.md
skills/architecture/SKILL.md
skills/brainstorming/SKILL.md
skills/canary/SKILL.md
skills/clone-patterns/SKILL.md
skills/code-review/SKILL.md
skills/compete/SKILL.md
skills/cost/SKILL.md
skills/demo/SKILL.md
skills/deploy/SKILL.md
skills/dispatching-parallel-agents/SKILL.md
skills/evals/SKILL.md
skills/executing-plans/SKILL.md
skills/finishing-a-development-branch/SKILL.md
skills/frontend-design/SKILL.md
skills/grow/SKILL.md
skills/guard/SKILL.md
skills/index-fix/SKILL.md
skills/investigate/SKILL.md
skills/launch/SKILL.md
skills/learn/SKILL.md
skills/onboard/SKILL.md
skills/pentest/SKILL.md
skills/perf-audit/SKILL.md
skills/receiving-code-review/SKILL.md
skills/release/SKILL.md
skills/requesting-code-review/SKILL.md
skills/rescue/SKILL.md
skills/retro/SKILL.md
skills/revise-claude-md/SKILL.md
skills/security-audit/SKILL.md
skills/seo-audit/SKILL.md
skills/seo-strategy/SKILL.md
skills/ship-gate/SKILL.md
skills/sprint/SKILL.md
skills/subagent-driven-development/SKILL.md
skills/systematic-debugging/SKILL.md
skills/test-driven-development/SKILL.md
skills/using-git-worktrees/SKILL.md
skills/using-ultraship/SKILL.md
skills/verification-before-completion/SKILL.md
skills/visual-diff/SKILL.md
skills/writing-plans/SKILL.md
skills/writing-skills/SKILL.md

Metadata

Files
0
Version
ed232cb
Hash
fea6b174
Indexed
2026-07-24 16:16

inicio - Wiki
Copyright © 2011-2026 iteam. Current version is 2.155.2. UTC+08:00, 2026-08-16 23:17
浙ICP备14020137号-1 $mapa de visitantes$