Salesforce Skill

Query and manage Salesforce CRM data via the Salesforce CLI (`sf`). Run SOQL/SOSL queries, inspect object schemas, create/update/delete records, bulk import/export, execute Apex, deploy metadata, and make raw REST API calls.

This skill appears to be what it says: a wrapper around the Salesforce CLI. Before installing, consider: 1) The CLI will require you to authenticate to a Salesforce org; those credentials (access tokens, refresh tokens, JWT key files, sfdx auth URLs) are sensitive — avoid exposing them to an agent you don't trust. 2) The skill can run destructive operations (create/update/delete, deploy, execute Apex). Only enable autonomous invocation if you want the agent to act on your org without manual approval. 3) Installing @salesforce/cli via npm may require elevated privileges and will run normal npm install scripts — if you prefer, install the official CLI yourself from developer.salesforce.com and keep credential files under your control. 4) If you need higher assurance, verify the npm package source and ownership before installing and prefer interactive (web) login flows over embedding keys in files where possible.

Type: OpenClaw Skill Name: salesforce Version: 1.0.0 The skill provides extensive and high-risk capabilities for interacting with Salesforce, including executing arbitrary Apex code, making raw REST/GraphQL API calls, and performing bulk data operations (create, update, delete) via the `sf` CLI. While the `SKILL.md` includes explicit guardrails instructing the AI agent to seek user confirmation for destructive actions, the inherent power of these commands (e.g., `sf apex run`, `sf api request rest`) means that a compromised agent or a malicious user could cause significant harm. There is no clear evidence of intentional malicious behavior or prompt injection designed to bypass these guardrails within the skill's definition itself, but the broad permissions and powerful actions classify it as suspicious.