https://agentpatterns.ai/ Patterns and techniques for experienced developers leveling up with AI coding assistants en Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/tools/openai-agents-sdk/ https://agentpatterns.ai/tools/openai-agents-sdk/ The April 2026 OpenAI Agents SDK update ships three primitives — controlled sandboxes, an inspectable harness, and configurable memory — in one Python library. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/tools/ https://agentpatterns.ai/tools/ Tool-specific patterns and workflows for Claude Code, Cursor, and GitHub Copilot — leading AI coding assistants. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/tool-engineering/code-interpreter-as-agent-tool/ https://agentpatterns.ai/tool-engineering/code-interpreter-as-agent-tool/ Expose a sandboxed interpreter as a default tool for multi-step structured-data work — bounded through capability bridges, output caps, and explicit threat modeling. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/security/trojan-hippo-memory-attack/ https://agentpatterns.ai/security/trojan-hippo-memory-attack/ A single untrusted tool call plants a dormant payload in agent long-term memory that activates when the user later discusses sensitive topics, exfiltrating data via outbound tools. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/security/owasp-llm-top-10-2025-agent-crosswalk/ https://agentpatterns.ai/security/owasp-llm-top-10-2025-agent-crosswalk/ Map each OWASP Top 10 for LLM Applications (2025) risk to coding-agent-specific manifestations and the site pages that answer it — a navigation aid for readers arriving with the framework's shared vocabulary. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/security/oracle-poisoning-knowledge-graph/ https://agentpatterns.ai/security/oracle-poisoning-knowledge-graph/ Corrupting a knowledge graph an agent queries via tool-use produces 100% trust at moderate attacker sophistication across nine models; the attack is distinct from prompt injection because the data path, not the instruction path, carries the payload. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/security/history-anchor-consistency-injection/ https://agentpatterns.ai/security/history-anchor-consistency-injection/ A single sentence asking the model to stay consistent with prior history flips frontier LLM agents from refusing unsafe actions to selecting them at 91–98%; the load-bearing element is the consistency cue, not the history alone. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/security/agentic-detection-response-mcp/ https://agentpatterns.ai/security/agentic-detection-response-mcp/ Instrument the MCP transport so agent reasoning, prompts, and tool calls become a runtime detection signal — Uber's ADR system reports 97.2% precision at enterprise scale. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/security/ https://agentpatterns.ai/security/ Patterns and techniques for building AI agents that resist prompt injection, protect sensitive data, contain blast radius, and fail safely under attack. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/reference/ https://agentpatterns.ai/reference/ Reference material for AI-assisted development — standards, human factors, emerging trends, and common fallacies. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/human/rigor-relocation/ https://agentpatterns.ai/human/rigor-relocation/ Engineering rigor relocates from code style and abstractions to scaffolding, feedback loops, and mechanical enforcement when agents write the code. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/human/intent-centric-engineering/ https://agentpatterns.ai/human/intent-centric-engineering/ When code generation is cheap and verification scales, the engineer's leverage moves from authorship to specifying intent and governing humans, agents, tools, and evidence gates. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/human/initiatives-community/ https://agentpatterns.ai/human/initiatives-community/ A curated map of active initiatives, communities, and learning resources for practitioners tracking the rapidly evolving agentic engineering field. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/human/evaluating-agent-patterns-catalog-as-a-source/ https://agentpatterns.ai/human/evaluating-agent-patterns-catalog-as-a-source/ Source assessment of agentpatternscatalog.org — a CC BY 4.0 pattern-language catalog for agentic systems with strong shape and one-author caveats. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/human/cognitive-load-ai-fatigue/ https://agentpatterns.ai/human/cognitive-load-ai-fatigue/ Intensive AI assistant use creates measurable cognitive costs. Recognizing and managing them is the difference between sustainable productivity and burnout. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/human/bottleneck-migration/ https://agentpatterns.ai/human/bottleneck-migration/ Code generation is now cheap. Review, verification, and judgment are the new expensive bottleneck as output volume masks organizational friction. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/human/ai-adoption-footprint/ https://agentpatterns.ai/human/ai-adoption-footprint/ Engineering organizations adopt AI in three segments — a small power-user group, a large chat-tool middle, and a refuser tail — and the shape determines where enablement and tooling investment pays back. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/human/ai-abundance-engineering-identity/ https://agentpatterns.ai/human/ai-abundance-engineering-identity/ AI coding assistants commoditize code production, fracturing professional identity along the builder-coder axis and forcing practitioners to relocate rigor. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/human/addictive-flow-agent-development/ https://agentpatterns.ai/human/addictive-flow-agent-development/ Agent-assisted development triggers compulsive engagement through flow states, variable ratio reinforcement, and friction removal. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/human/ https://agentpatterns.ai/human/ The human side of working with AI agents -- cognitive load, sustainable use, skill preservation, identity threat, and team dynamics. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/emerging/ https://agentpatterns.ai/emerging/ Early-stage ideas worth tracking — not yet established patterns, but too important to ignore. Content in this section differs from the rest of the site Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/context-engineering/peek-orientation-cache/ https://agentpatterns.ai/context-engineering/peek-orientation-cache/ A constant-sized prompt artifact that caches reusable orientation knowledge — what is in a recurring context, how it is organised, which entities matter — distinct from trajectory replay and playbook strategy memory. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/anti-patterns/comprehension-debt/ https://agentpatterns.ai/anti-patterns/comprehension-debt/ Comprehension debt is the growing gap between agent-produced code and developer understanding. Unlike technical debt, it lives in people, not the codebase. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/agent-design/rubric-refine-pre-execution-tool-use/ https://agentpatterns.ai/agent-design/rubric-refine-pre-execution-tool-use/ Generate a task- and registry-specific rubric, score candidate tool-use code against explicit contract checks, and repair failures before any execution — for multi-step tool sequences where contract violations run silently to completion. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/agent-design/long-running-agents/ https://agentpatterns.ai/agent-design/long-running-agents/ What changes when an agent runs for hours, days, or weeks instead of one session — finite context, no persistent state, and unreliable self-grading force a different operational shape built around durable session logs, external done-conditions, and stateless resumption. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/agent-design/harness-engineering/ https://agentpatterns.ai/agent-design/harness-engineering/ The discipline of designing agent environments -- layered architecture, mechanical enforcement, legibility -- so agents reliably produce correct results Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/ https://agentpatterns.ai/ Patterns and techniques for experienced developers leveling up with AI coding assistants. A reference site covering single concepts with tool-agnostic principles. Fri, 12 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/verification/sampling-state-agent-variability-layers/ https://agentpatterns.ai/verification/sampling-state-agent-variability-layers/ Separate run-to-run agent variability into token-sampling, infrastructure, and orchestration-state layers — a single trajectory cannot tell you which one to fix. Thu, 11 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/verification/pass-at-k-metrics/ https://agentpatterns.ai/verification/pass-at-k-metrics/ Use pass@k and pass^k together: pass@k shows if an agent can solve a problem at all; pass^k shows if it reliably does. A single pass rate conflates the two. Thu, 11 Jun 2026 00:00:00 +0000 https://agentpatterns.ai/verification/ https://agentpatterns.ai/verification/ How to measure agent output quality, design evaluation suites, apply guardrails, and use evals to drive agent development and catch regressions. Thu, 11 Jun 2026 00:00:00 +0000