Spend Pulse

Proactive spending alerts via Plaid. Track credit card spending against a monthly budget with pace-based alerts.

This skill appears to implement a Plaid-based spending-alert CLI, but it does not include code in the bundle and asks you to install an external npm package or clone a GitHub repo. Before installing or handing it Plaid credentials: 1) Verify the npm package and GitHub repo (owner, stars, recent commits, package versions, publisher identity). 2) Inspect the package source code (or ask for an audited release) to confirm it only uses Plaid and writes config under ~/.spend-pulse and the macOS Keychain as claimed. 3) Confirm you are comfortable installing third‑party npm packages (they can run arbitrary code at install/run time). 4) Note the instructions target macOS Keychain — ensure this matches your OS or that the tool provides a secure alternative. 5) If you allow the agent to invoke the skill autonomously, remember it could run the CLI and access stored financial data; only enable autonomous use if you trust the package source. If you want higher assurance, request a skill bundle that includes audited code or a clear install spec from a verified release (GitHub release or npm package with known publisher) and explicit metadata linking the registry entry to the repository.

Type: OpenClaw Skill Name: spend-pulse Version: 0.1.2 The OpenClaw AgentSkills skill bundle 'spend-pulse' appears benign. The `SKILL.md` file describes a financial tracking tool that integrates with Plaid, requiring access to financial data and API credentials. It explicitly states that credentials are stored securely in macOS Keychain. The skill generates a chart image (`spend-pulse chart`) and instructs the agent to attach it to messages, which is a legitimate use of agent capabilities for visual communication. All described actions and instructions for the agent are directly aligned with the stated purpose of providing proactive spending alerts, and there is no evidence of intentional harmful behavior, data exfiltration, malicious execution, persistence mechanisms, or prompt injection attempts to subvert the agent's core function.