Results for “agent-tool-abuse”

Apply least-privilege tool allowlisting, identity binding, and human-in-the-loop controls for agent tool calls.

Add governance, safety, and trust controls to AI agent systems with policy enforcement, intent classification, and audit trails.

Creates or modifies agent tools for OpenAI Agents SDK, covering search, analysis, API integration, and database query patterns with caching and metrics.

Creates or modifies OpenAI Agents SDK tools with Zod parameters, caching, metrics, and agent registration, following established patterns and best practices.

Debug agent execution by querying trace and metric tables, analyzing tool calls, durations, errors, and performance trends.

Design multi-agent system architectures, generate tool schemas for Anthropic and OpenAI formats, and evaluate execution logs for cost, latency, and failure bottlenecks.

Discovers, installs, and manages AI agent skills from agentskill.sh, including searching, installing mid-session, scanning for security issues, and providing feedback.

Scaffold and develop AI agents using OpenAI Agents SDK patterns, covering agent definition, tools, guardrails, handoffs, context, RAG pipelines, streaming, API routes, testing, and debugging.

Investigates slow responses, tool failures, and guardrail rejections by querying agent trace logs and performance metrics.

Design tool descriptions, schemas, and response formats that agents can reliably route on, including MCP server design, tool-set consolidation, and naming conventions.

Compare coding agents head-to-head on reproducible tasks with pass rate, cost, time, and consistency metrics.