PRISM OS SDK

Financial data SDK for AI Agents. 218+ read-only endpoints for market data, prices, fundamentals. Built for Cursor, Claude, OpenClaw. Data retrieval only.

This package mostly looks like a comprehensive market-data SDK and correctly asks only for PRISM_API_KEY. The red flag is inconsistency: SKILL.md promises read-only behavior, but the repository and examples include 'execute' modules, simulate/execute flows, and references to swap execution that could accept a signer and produce transaction hashes. Before installing or enabling autonomous use: 1) Inspect src/modules/execute (and any onchain/dex files) to confirm they are truly simulation-only and do not send transactions or accept private keys. 2) Review package.json and package-lock for third-party dependencies and any postinstall scripts. 3) Verify the npm package and GitHub repo (authenticity and recent commits) match the published package. 4) Do not provide private keys or wallet mnemonics to the skill; if you need execution, prefer explicit signer objects you control in a sandbox. 5) If you want a safer setup, restrict autonomous invocation (disable model invocation for this skill) until you audit the execute functionality. If you want, I can scan specific source files (e.g., src/modules/execute/index.ts and any files that reference signer/wallet or external webhook URLs) and summarize exactly where execution or outbound side-effects occur.

Type: OpenClaw Skill Name: prism-finance-os Version: 0.1.2 The OpenClaw AgentSkills bundle provides a powerful SDK for financial operations, including autonomous trading across crypto, DeFi, stocks, forex, and prediction markets. While the code itself appears to be a legitimate tool for its stated purpose, the `SKILL.md` documentation contains misleading statements, claiming 'Data retrieval only' and 'No trading execution', which directly contradict the extensive execution capabilities demonstrated in the `src/modules/execute/index.ts` module and the provided example agents. This discrepancy, combined with the inherent high-risk nature of autonomous financial transactions (e.g., market buys/sells, arbitrage, portfolio rebalancing), raises significant transparency concerns and could lead to an agent or user underestimating the skill's true capabilities and potential for financial loss if misused.