Claude api
Repository-first deterministic migration and atomic Skill source for Claude Code and Codex.
npx -y skills add JasonxzWen/harness-hub --skill claude-apiAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
What its author says it does
Copied from the file, not written here
Load when a task needs Claude API or Anthropic SDK build, debug, migration, or tuning work; skip provider-neutral, OpenAI, or non-Anthropic model work.
The file declares its own license as Complete terms in LICENSE.txt. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.
SKILL.md
2.3 KB, as published. Nobody here has run it
Claude API
Use this skill for provider-specific Claude API and Anthropic SDK work. Keep it explicit: if the target file is provider-neutral or already uses another provider, stop and ask whether the user wants an Anthropic-specific implementation.
Workflow
- Detect the project language, runtime, and existing SDK usage.
- Read
references/live-sources.md, then verify current Anthropic docs or SDK source before relying on API names, model IDs, headers, beta flags, or migration details that may have changed. - Prefer the official Anthropic SDK for the project language. Use raw HTTP only when the user asks for REST/cURL or no suitable SDK exists.
- Keep secrets in environment variables or the target repo's existing secret manager. Never place API keys, tokens, or credentials in prompts, committed files, examples, or logs.
- Preserve existing provider boundaries. Do not introduce Anthropic SDK calls into OpenAI, generic, or multi-provider files unless the accepted change is to add an Anthropic lane.
- Run the nearest compile, typecheck, unit, or smoke check after changing code.
Implementation Checks
- Confirm exact package names, imports, client constructors, method names, request fields, streaming helpers, and response shapes from current docs or SDK source.
- Treat model names, pricing, context windows, rate limits, prompt caching behavior, and migration guidance as live facts.
- Make cost, latency, timeout, and retry tradeoffs explicit when changing generation behavior.
- Add tests or fixtures around request construction, response parsing, and error handling where the project already has a test surface.
Boundaries
- Use
documentation-lookupfor general API/library questions that are not Anthropic-specific. - Use
mcp-builderfor MCP server design or tool schemas. - Use
security-reviewwhen auth, secret handling, sandboxing, or untrusted tool execution is central.