Development
#agent
check-sdk-updates
Check whether Claude Code, Claude Agent SDK, and Codex SDK have new versions on npm. Use when the user asks to check for SDK/CLI updates, e.g. "check sdk updates", "check claude code update", "有沒有更新", "檢查 SDK 版本".
DeepseekModel
Curated skill
Quality Excellent · 90
v1.0.0
Get
https://deepseekmodel.com/api/download.php?id=tony1223-better-agent-terminal-claude-skills-check-sdk-updates-skill-md&format=skill
Download .skill
Standard format with system_prompt and model_config, ready for any agent framework
The actual content of the system_prompt field in the .skill file.
name check-sdk-updates description Check whether Claude Code, Claude Agent SDK, and Codex SDK have new versions on npm. Use when the user asks to check for SDK/CLI updates, e.g. "check sdk updates", "check claude code update", "有沒有更新", "檢查 SDK 版本". Check SDK Updates Compare the versions of @anthropic-ai/claude-code , @anthropic-ai/claude-agent-sdk , and @openai/codex-sdk declared in package.json against the latest published version on npm. Steps Read the dependencies block of package.json and extract the current declared range for each of: @anthropic-ai/claude-code @anthropic-ai/claude-agent-sdk @openai/codex-sdk Run the following command (single Bash call) to fetch latest versions in parallel: npm view @anthropic-ai/claude-code version & \ npm view @anthropic-ai/claude-agent-sdk version & \ npm view @openai/codex-sdk version & \ wait Map the three lines of output back to the three packages in the order requested. For each package, compare the declared version (strip leading ^ / ~ ) with the latest. Report a compact table: Package Current Latest Status @anthropic-ai/claude-code 2.1.119 2.1.120 ⬆ update @anthropic-ai/claude-agent-sdk 0.2.119 0.2.119 ✓ up to date @openai/codex-sdk 0.125.0 0.126.0 ⬆ update For each outdated package, analyze the gap between current and latest so the user can decide how to implement the update. For each one: a. Get the list of versions in between with: npm view <pkg> versions --json Filter to versions strictly greater than current and ≤ latest. Note the count and span (patch / minor / major bumps). b. Try to fetch release notes / changelog. Try in this order, stop at the first that works: npm view <pkg> repository.url → if it points to a GitHub repo, use gh release list -R <owner/repo> --limit 30 and gh release view <tag> -R <owner/repo> for the relevant tags. Look for a CHANGELOG.md inside node_modules/<pkg>/ (already-installed copy reflects the current version, but its history section may still cover newer versions if upstream backfills). Fall back to: npm view <pkg>@<latest> --json and read the description / homepage fields, then WebFetch the homepage's changelog/releases page. c. Summarise per package in 2–4 bullets: Number of versions skipped (e.g. "5 patch releases, 1 minor") Any breaking changes or notable API/CLI changes Any items that touch the parts of our codebase that consume this SDK (e.g. electron/claude-agent-manager.ts , electron/codex-agent-manager.ts , electron/openai-agent-manager.ts ) Risk assessment: low / medium / high, with one-line justification After the per-package analysis, suggest the exact npm install command to bump them, e.g.: npm install @anthropic-ai/claude-code@latest @openai/codex-sdk@latest Group low-risk bumps together; call out any high-risk bump as needing a separate, careful update. Do NOT run the install — only suggest it. Wait for the user's confirmation before installing. Notes Use npm view (not npm outdated ) so the check works without touching the lockfile or node_modules. If npm view fails (offline / registry error), report which package failed and stop. Keep the output short — table + suggested install line is enough.
Keywords that activate this skill. Click one to copy it.
This skill does not provide trigger words.
The downloaded .skill package contains the following fields.
| Field | Description |
|---|---|
| format | Format tag (skill/v1) |
| skill_id | Unique skill ID |
| name | Skill name |
| version | Version |
| description | Description |
| category | Categories (array) |
| trigger_words | Trigger words |
| tags | Tags |
| source | Source |
| source_url | Source URL (this page) |
| exported_at | Exported at (set per download) |
| system_prompt | System prompt body |
| model_config | Model config: provider / model / temperature / max_tokens / top_p |
| examples | Examples |
| install_guide | Import guide for Coze / Dify / Claude / custom frameworks |
The same skill can be exported in different platform formats.