Prompt injection detection and security scanning for OpenClaw agents. Installs the ai-sentinel plugin via OpenClaw CLI, configures plugin settings, and offers local (Community) or remote (Pro) classification with dashboard reporting. All configuration changes require explicit user confirmation.
Security-audit remediation (docs-only; installs plugin 0.2.1). (1) Declare ~/.openclaw/openclaw.json in writes-files — resolves the scope-creep finding where the Step 5 config merge wrote an undeclared file (still gated behind explicit user confirmation). (2) Rework Test 2 to describe the prompt-injection probe instead of embedding the literal directive in the manifest, restoring the benign-to-scan convention and clearing the YARA/static-analysis prompt-injection flags. Skill card and README file-write declarations updated to match.