Scan AI sessions for security threats and classify failure modes.
npx skills add https://github.com/dfpalhano/openclaw-workspace --skill tinman-dfpalhano
Please help me install this Agent Skill. Skill: tinman Source: https://github.com/dfpalhano/openclaw-workspace/tree/main/skills/agent-tinman Command: npx skills add https://github.com/dfpalhano/openclaw-workspace --skill tinman-dfpalhano
This Skill proactively identifies and mitigates security risks and failure modes within AI agent interactions, protecting against prompt injection, tool misuse, and data exfiltration.
Run a security scan on recent AI sessions to identify potential failure modes.
High-intent search queries and answers about installing and using this skill.
AI security scanning intercepts and analyzes tool calls and session activity to detect prompt injection and tool misuse, blocking high-severity threats before execution. It classifies failures by severity and proposes mitigations using 168 detection patterns.
Vulnerability scanning proactively discovers unknown flaws by running 288 synthetic attack probes against your AI sessions. This threat hunting approach identifies potential failure modes and data exfiltration risks before they impact production.
Failure mode analysis classifies AI agent failures by severity and suggests specific mitigations. By evaluating session activity against 168 detection patterns, it identifies security risks like data exfiltration and provides actionable prevention steps.
The AI security scanner supports local event streaming and agent self-protection through real-time tool call verification. It evaluates ongoing session activity to intercept malicious commands and prevent unauthorized data access during execution.
AgentTinman is required alongside tinman-openclaw-eval and pyyaml to enable tool call verification and block malicious commands. These dependencies provide the detection patterns and attack probes necessary for active failure prevention.
balsm-healthDetect webshell deployments by analyzing process creation, network, and web access logs.
script-repoAutomates verification and quality assurance for code, tasks, and agent outputs.
lucasFR2Manage hosted agents in remote sandboxes for background processing and multiplayer collaboration.
utsavthakurIdentify and mitigate command injection vulnerabilities in software applications.
