0xcipher0/cosin/SKILL.md
Use this skill when an agent needs to operate the `cosin` CLI from the terminal. `cosin` accepts only relative paths, lists available skills through the `skills` subcommand, can call COS API paths directly, and can call `/cos/...` skills by turning them into x402 pay-and-call requests to the COS backend. This skill is for using the CLI, not for editing the cosin source code.
npx skillsauth add openclaw/skills cosinInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Use cosin to call the COS API and COS-backed skills from the terminal.
cosin now accepts only relative paths.
There are three request styles:
skills
Calls https://skills.bankofuniverse.org/skills directly and prints the upstream body directly.
Normal COS API paths such as /v1/me
These are sent directly to the COS API host.
Skill paths under /cos/...
These are converted internally into x402 pay-and-call requests:
SKILLS_BASE_URLPOST /agent/pay-and-call on the COS API hostUsers should not pass absolute URLs to the CLI anymore.
Collect these inputs before running the CLI:
--key/--json--header or -H--base-url override for normal COS API callsAsk for the token if the user has not provided one. Treat it as sensitive.
--key value as sensitiveUse one of these command shapes:
cosin --key <token> <METHOD> <PATH> [--json '<json>'] [--header 'Name: value'] [--base-url <url>]
cosin --key <token> agent [status|me] [--base-url <url>]
cosin --key <token> skills
Important flags:
--key <token> for the required bearer token--json <json> for an optional JSON request body--header 'Name: value' or -H 'Name: value' for repeatable custom headers--base-url <url> to override the default COS API base URL for direct API calls--version or -v to print the installed CLI version--help or -h to print usageSubcommand notes:
skills does not accept --jsonskills does not accept custom headersagent does not accept --jsonagent does not accept custom headersUse this to discover available skills:
cosin --key <token> skills
Expected upstream skills include:
/cos/crypto/chainlink/random
Returns a random value from the Chainlink-based skill endpoint./cos/crypto/price/:symbol
Returns the latest price for a supported token symbol.Supported symbols for /cos/crypto/price/:symbol:
BTCETHHYPESOLTRXUSDTUSDCUse normal API paths to call COS directly:
cosin --key <token> GET /v1/me
cosin --key <token> POST /v1/orders --json '{"symbol":"BTCUSDT"}'
/cos/...Use /cos/... when you want to call a skill through COS:
cosin --key <token> GET /cos/crypto/chainlink/random
cosin --key <token> GET /cos/crypto/price/BTC
Internally, the CLI turns those into x402 pay-and-call requests to the COS backend.
Use agent, agent status, or agent me as a convenience alias for GET /agent/me.
cosin --key <token> agent
cosin --key <token> agent status
cosin --key <token> agent me
Do not combine agent with --json or custom headers.
/skills, not GET /skills--json is valid JSONName: value/cos/crypto/chainlink/random, only use GET/cos/crypto/price/:symbol, only use GET and only the supported symbolsExpect the CLI to:
/cos/... skill calls, and agentskills subcommand<empty response body> for empty responsesUse the local build when working inside this repository:
bun run build
./dist/index.js --key <token> agent
./dist/index.js --key <token> skills
./dist/index.js --key <token> GET /v1/me
./dist/index.js --key <token> GET /cos/crypto/chainlink/random
Use the published command shape locally when that is more convenient:
bun x cosin --key <token> agent
bun x cosin --key <token> skills
bun x cosin --key <token> GET /v1/me
bun x cosin --key <token> GET /cos/crypto/price/BTC
tools
Use when the user wants to connect to, test, or use the McDonalds service at mcp.mcd.cn, including checking authentication, probing MCP endpoints, listing tools, or calling McDonalds MCP tools through a reusable local CLI.
development
Web scraping platform — Twitter/X data, Vinted marketplace, and general web scraping API
development
SlowMist AI Agent Security Review — comprehensive security framework for skills, repositories, URLs, on-chain addresses, and products (Claude Code version)
data-ai
去除中文文本中的 AI 写作痕迹,使其读起来自然。基于维基百科 AI 写作特征指南,检测 24 种 AI 模式。触发词:humanizer-cn、去除 AI 痕迹、去除 AI 写作痕迹、中文文本人性化。