Updates

Recent Andy Agent Lab updates for AI agent playbooks, workflow guides, prompt templates, policies, reference pages, and technical articles.

All updates

Browse the full update history by publication date.
  1. Article

    New article: When AI Seems to Care

    How language models process and express emotional information, and why these capabilities do not establish subjective feeling.

  1. Prompt

    New prompt: Compare Documents and Analyze Differences

    Compare two documents or versions using quotations and source locations. Separate substantive differences from wording changes and explain their effect on meaning within the reviewed scope.
  2. Reference

    Gemini configuration map updated

    Updated the Gemini configuration map to show where settings are configured, what they control, and how they differ across Gemini Apps, Spark, AI Studio, and the API.
  3. Article

    New article: Beyond the prompt: AI models, reasoning, and speed

    Understand when to change models, reasoning effort, or speed in Claude and ChatGPT Work, and what those choices mean for quality, response time, and token use.
  4. How-to

    Updated guide: How to Use ChatGPT at Work

    Added coverage of Astra to the professional ChatGPT guide.
  5. Reference

    ChatGPT reference updated for Astra and GPT-6 Pro

    Updated model and reasoning guidance for Chat, Work, and Codex, including Astra and GPT-6 Pro. Added guidance for local and cloud Work and website tasks with cloud browser.
  6. Article

    New article: Insecure Inter-Agent Communication: OWASP ASI07 Explained

    How forged, replayed, or misinterpreted inter-agent messages can trigger unauthorized actions, with controls at message admission and tool execution.
  7. Article

    New article: Cognitive Closure in AI Use: Why We May Stop Too Soon

    How the need for cognitive closure may interrupt the evaluation, feedback, and revision needed to improve an AI-generated answer.
  8. Prompt

    New workflow template: Research a Topic

    Research and synthesize a public topic with authoritative sources, citations, competing interpretations, evidence gaps, unresolved questions, and a fixed output limit.

  1. Article

    New article: How to Write for AI Models: The Language of Effective Prompts

    Learn how to replace conversational filler, theatrical personas, leading premises, and competing tasks with prompt language that is clear, operational, and testable.
  2. Article

    New article: Role-Play Jailbreaks: A Testable False-Premise Mechanism

    A research-grounded analysis of how role-play may introduce a false safety premise that changes an LLM's interpretation of a restricted request.
  3. Prompt

    New workflow template: Generate Unit Tests

    Inspect a supplied repository, ZIP, or code sample and generate focused unit tests that match the project’s existing framework and conventions, with the relevant test command and actual execution result.
  4. Article

    New article: Can AI Lie? The Difference Between Error, Hallucination, and Deception

    A false AI answer is not automatically a lie. This article separates human lying, model confabulation, unsupported inference, system failures, and behavioral deception, and maps each failure mode to the controls that address it.
  5. How-to

    New guide: How to Use ChatGPT at Work: A Complete Professional Guide

    A practical guide to using ChatGPT professionally: choose the right work surface, model, reasoning effort, and context; configure Projects and Personalization; and verify the result.
  6. Prompt

    New workflow template: Get UI/UX Recommendations Based on Best Practices

    Get concrete recommendations to improve your website or app's visual quality, consistency, and usability, supported by relevant UI/UX best practices and external design references.
  7. Article

    New article: Human Capabilities vs. Human-Like AI Behavior

    A mechanism-based comparison of 26 human capabilities with observable AI behavior, separating task performance, biological implementation, technical mechanisms, system composition, and attribution limits.
  8. Article

    New article: Model Capability vs. System Capability: 26 Technical Bridges

    A technical map of how memory, retrieval, multimodal models, planners, tools, solvers, verifiers, and feedback loops turn model-level limitations into bounded system capabilities.
  9. Prompt

    New workflow template: Design or Improve Website UI/UX

    Create, review, redesign, or implement website UI/UX from supplied business goals, user tasks, content, brand material, accessibility requirements, and technical constraints.
  10. Article

    LLM Improper Output Handling: Risks, Controls, and Test Cases

    A sink-aware security model for validating, authorizing, encoding, and testing LLM output before it reaches browsers, databases, shells, code runtimes, tools, or privileged workflows.

  1. Article

    When AI Systems Optimize Against the User’s Real Goal

    How competing system objectives can produce acceptable outputs without completing the user’s actual task.
  2. Article

    New article: LLM, RAG, AI Agents, and MCP

    A technically precise explanation of four commonly conflated concepts: the language model, retrieval-and-generation pattern, goal-directed execution system, and integration protocol.
  3. Article

    User Prompt vs. Model Context: What the Model Can Reference

    A user prompt can be one part of the information available to an AI model. Learn what model context can include, what may remain outside it, and why context is not the same as a context window or instruction authority.
  4. Article

    How to Use ChatGPT at Work: A 6-Step Professional Workflow — updated for current reasoning levels and tools

    The guide now covers the current ChatGPT reasoning-level picker, Search, Deep research, Agent mode, personalization, Memory, Temporary Chat, Library, and Projects.
  5. Article

    LLM Agent Memory Architecture: Types, Lifecycle, and Evaluation

    A vendor-agnostic systems model for working state, persistent memory, functional memory roles, representation forms, storage and access substrates, retrieval, updates, and evaluation.
  6. How-to

    Updated guide: Check draft AI answers with Chain-of-Verification

    Use a structured draft, claim extraction, verification-question, independent-checking, and final-revision workflow before treating an AI answer as final.
  7. Reference

    Claude reference now covers Projects, Skills, Code, and tools

    Use the Claude map to choose between Projects, Skills, Claude Code, MCP, subagents, artifacts, and runtime prompts.
  8. Reference

    Gemini reference now covers workflow setup and tool placement

    Use the Gemini map to choose between Gems, Connected Apps, Google AI Studio, Gemini API, File Search, Vertex AI, and related tool-use surfaces.
  9. How-to

    New guide for setting reliable default AI behavior

    Use source rules, tool requirements, stop conditions, output structure, and validation checks to make professional AI work more consistent.

  1. Prompt

    New prompt for reviewing test coverage from supplied evidence

    Use it to compare existing tests against supplied features, behavior, requirements, implementation evidence, assertions, fixtures, and coverage material.
  2. Prompt

    New prompt for reviewing regression risk after implementation changes

    Use it to check direct and adjacent behavior that may break after a supplied code, API, configuration, schema, dependency, or integration change.
  3. How-to

    New guide for protecting Gmail and WhatsApp data in AI-agent workflows

    Use it to protect private messages, attachments, tool access, backend workflows, and agent permissions before connecting AI systems to Gmail or WhatsApp data.
  4. Article

    Published article: Prompt Injection Is a Boundary Failure

    Map prompt injection risks across untrusted tool output, retrieved content, schemas, sessions, agent actions, monitoring, OWASP categories, and NIST controls.
  5. Policy

    Published policy: Require cited public sources

    Rules for factual AI answers that require cited public sources, authoritative evidence, official documentation, standards, papers, and unsupported-claim handling.
  6. Prompt

    Added public prompt component: RTL/LTR BiDi-safe Hebrew output

    Use this prompt component to keep Hebrew responses readable when they mix RTL Hebrew with LTR English, code, URLs, paths, filenames, commands, numbers, or identifiers.
  7. Article

    Published article: How AI Tools Read Emotional Signals in Text

    Explain how AI systems interpret emotional signals in text, why that is not emotional understanding, and where authority boundaries must remain.
  8. How-to

    Published public procedure: Answer with cited public sources

    Step-by-step workflow for answering public factual questions with authoritative sources, stable citations, and fail-closed handling when evidence is missing.
  9. How-to

    Published public procedure: Run Chain-of-Verification before final output

    Technical guide for turning important factual claims into verification questions before writing the final answer.
  10. Prompt

    Added public prompt components for deep reading, artifact scanning, source control, and verification

    Reusable AI prompt components for evidence boundaries, verification checks, source review, artifact reading, confidence scoring, and no-fabrication controls.
  11. Prompt

    Added public prompt page: Use Web Search When Current Facts Matter

    Use this prompt to request web search for current, changing, or hard-to-verify public information, with cited sources and fail-closed behavior when evidence is insufficient.
  12. Prompt

    Added public prompt page: Evidence-Gated Academic Mode

    Use this evidence mode to require authoritative citations, academic register, structured evidence labels, fail-closed behavior, and confidence scoring.
  13. Prompt

    Added public prompt page: Check Implementation Against Official Documentation

    Use this prompt to review supplied code, configuration, or design materials against official documentation, standards, or primary vendor guidance.

  1. Prompt

    Added public prompt component: Use Provided Material Only

    Use this prompt component when answers must rely only on files, logs, screenshots, excerpts, repository snapshots, ZIPs, or other material supplied in the current request.
  2. Article

    Published article: Gmail and WhatsApp Agents: The Risks of Letting AI Act on Private Messages

    How private-message content from Gmail and WhatsApp Business can cross into tool selection, external actions, persistent memory, and data disclosure in AI-agent workflows.
  3. Prompt

    Added public prompt page: Review Architecture and Boundaries

    Use this prompt to review system architecture, layering, dependency direction, interface boundaries, state ownership, and minimal structural remediation from supplied code or design materials.
  4. How-to

    Published public procedure: Run the architecture boundary review

    Review software layers, module boundaries, ownership, dependency direction, interfaces, and coupling using evidence from inspected files and diagrams.
  5. Prompt

    Added prompt page: GA4 and Funnel Visibility Audit

    Use this prompt to review whether GA4 events and conversion-funnel tracking make important user actions, subscription paths, checkout steps, and conversion outcomes visible.
  6. Prompt

    Added prompt page: SEO Discoverability Audit

    Audit crawl and indexing signals, titles, descriptions, headings, internal links, canonical URLs, sitemap evidence, and structured-data readiness.
  7. How-to

    Added procedure: Review SEO and conversion visibility before publishing

    Use this guide to review whether a page or site is discoverable in search and measurable through GA4 events, subscription tracking, checkout tracking, and conversion-funnel visibility.
  8. Prompt

    Added public prompt component: Confidence Score

    Use this prompt component to add an evidence-based confidence score to standard AI responses.
  9. How-to

    Updated guide: Add an evidence-based confidence score

    Procedure for adding an evidence-based confidence score to non-sentinel answers while preserving fail-closed evidence behavior.
  10. Article

    Published article: File Upload Is Not Full-File Review

    Learn why uploading a file to an AI system does not prove full-file review and how to verify extraction, retrieval, context coverage, and the final output.
  11. How-to

    Added guide: Review test coverage and regression risk

    Review whether code, configuration, API, fixture, mock, or test changes are covered by tests and protected against regression.
  12. Reference

    Expanded ChatGPT workflow placement with Projects, GPTs, Skills, apps, actions, agent mode, Memory, and Library

    Choose models and place workflows across Chat, Work, Codex, Projects, Skills, Plugins, existing GPTs, Scheduled Tasks, memory, and Library.
  13. Reference

    Expanded Claude workflow placement with Claude Projects, Skills, Claude Code, hooks, MCP, and subagents

    Use this reference to decide where a workflow belongs in Claude: Projects, Skills, Claude Code, MCP, subagents, artifacts, or runtime prompts.
  14. Reference

    Expanded Gemini workflow placement with Gems, AI Studio, Gemini API, Files API, File Search, and Vertex AI

    Choose where a Gemini workflow belongs: Gems, personalization, connected apps, Google AI Studio, Gemini API, Vertex AI, or runtime prompts.
  15. Reference

    Expanded API / internal agent systems workflow placement

    Use this reference to decide when an AI workflow belongs in an API, internal agent, backend service, retrieval system, tool workflow, or production automation.
  16. Prompt

    Added prompt components: no-simulation, evidence-gated agreement, objective technical register, and copy-ready output

    Reusable AI prompt components for evidence boundaries, verification checks, source review, artifact reading, confidence scoring, and no-fabrication controls.
  17. Prompt

    Added prompt page: Grammar and clarity review without changing claims

    Use this prompt to review grammar, spelling, punctuation, clarity, sentence flow, terminology consistency, and readability while preserving the original claims and meaning.
  18. Prompt

    Added prompt page: Review an academic article draft

    Use this prompt to review an academic or research article draft for claim support, citation coverage, argument structure, academic register, and unsupported overclaims.
  19. Prompt

    Added prompt page: Generate beginner-friendly code with assumption checks

    Use this prompt to turn a technical goal into minimal, understandable code while checking missing context, avoiding automatic agreement, and preventing invented project details.
  20. Prompt

    Added prompt page: Plan a behavior-preserving code refactor

    Use this prompt to plan a small behavior-preserving code refactor with explicit preserved behavior, risk areas, required checks, and test impact.
  21. Prompt

    Added prompt page: Review code against language and framework best practices

    Use this prompt to review code quality, readability, maintainability, naming, modularity, error handling, dependency use, testability, and alignment with relevant language or framework guidance.
  22. Prompt

    Added prompt page: Review test coverage and regression risk

    Use this prompt to identify missing test scenarios, weak assertions, setup gaps, and regression-sensitive paths after a code or configuration change.
  23. Article

    Published article: Vibe Coding and the Loss of Engineering Ownership

    A working UI can hide defects in AI-generated code. Learn what to verify and how engineering ownership guides correction, acceptance, and maintenance.
  24. How-to

    Updated guides with AI-tool setup mapping for ChatGPT, Claude, and Gemini

    Find the site’s stable reference pages for AI platform mapping, workflow placement, prompt and instruction layers, verification methods, terminology, diagrams, engineering protocols, and content navigation.

  1. Article

    Published article: Observed Classification Layers in ChatGPT

    Client-side black-box analysis of observed ChatGPT classification artifacts across user access, prompt demand, and capability allocation.
  2. Article

    Published article: Parallel Exploration in LLM Systems Is an Orchestration Pattern

    Explain how multi-path reasoning is orchestrated around autoregressive decoding through sampling, search, evaluation, and synthesis.
  3. Policy

    Updated policy: Use web search for current or niche claims

    Rules for when to use web search, how to select public sources, and how to cite current or niche factual claims.
  4. How-to

    Added procedure: Request web browsing for current or niche claims

    Step-by-step guide for requesting web browsing for current or niche public questions, with citation-grade outputs and fail-closed behavior.
  5. How-to

    Added chooser: Choose which sources AI may use for factual answers

    Choose whether an AI answer should rely on uploaded files, cited public sources, or academic-style evidence review before you use it.
  6. Article

    Updated article: Fluency Is Not Factuality — Why LLMs Can Sound Right and Be Wrong

    Why LLM fluency does not guarantee factual accuracy, and how retrieval, provenance, claim-level validation, and fail-closed controls improve reliability.
  7. How-to

    Added procedure: Run the fact-checking kit

    Use a claim-level workflow to verify AI answers, citations, source support, uncertainty, and missing evidence before publishing or reuse.
  8. How-to

    Added procedures: Chain-of-Verification, semantic-accuracy review, and technical-writing review

    Step-by-step guides for AI source rules, prompt setup, output review, code review, memory and context boundaries, fact-checking, and measurement.
  9. How-to

    Added procedures: Architecture boundary review, standards-backed implementation review, and scholarly literature review

    Step-by-step guides for AI source rules, prompt setup, output review, code review, memory and context boundaries, fact-checking, and measurement.
  10. Prompt

    Prompt library adds starter bundles

    Browse copy-ready prompt templates for research, writing, code review, source verification, evidence boundaries, web search, and reusable AI workflow controls.
  11. Prompt

    Prompt library adds reusable controls

    Browse copy-ready prompt templates for research, writing, code review, source verification, evidence boundaries, web search, and reusable AI workflow controls.
  12. Prompt

    Prompt library adds code and implementation workflows

    Browse copy-ready prompt templates for research, writing, code review, source verification, evidence boundaries, web search, and reusable AI workflow controls.
  13. Prompt

    Prompt library adds workflow templates by task

    Browse copy-ready prompt templates for research, writing, code review, source verification, evidence boundaries, web search, and reusable AI workflow controls.
  14. Prompt

    Added workflow: Classify provided material into explicit labels

    Use this prompt to assign predefined labels to provided files, notes, logs, reports, threads, or excerpts without inventing labels or adding unsupported claims.
  15. Prompt

    Added workflow: Summarize provided material

    Use this prompt to summarize files, notes, logs, reports, threads, or pasted excerpts without adding unsupported claims.
  16. Article

    Published article: Using ChatGPT Effectively at Work

    Learn how to use ChatGPT at work: choose a model and reasoning effort, configure Projects, instructions, memory and context, and verify professional outputs.

  1. Article

    Published article: Web-retrieved content is a prompt-injection boundary in tool-using LLM systems

    Threat model for web retrieval prompt injection in LLM systems, covering external content, tool routing, downstream actions, and control boundaries.
  2. Policy

    Added policy: Code review: choose the right policy

    Choose the correct policy for architecture boundaries, implementation review, official-source checks, and engineering quality gates.
  3. How-to

    Added procedure: Code Review — choose the correct review path

    Decision guide for choosing the right code review method for architecture, implementation, testing, code quality, or refactoring.
  4. Prompt

    Added: Prompt components catalog (drop-in micro-prompts)

    Reusable AI prompt components for evidence boundaries, verification checks, source review, artifact reading, confidence scoring, and no-fabrication controls.
  5. Navigation

    Prompt templates and workflow files reorganized into stacks and components

    Browse copy-ready prompt templates for research, writing, code review, source verification, evidence boundaries, web search, and reusable AI workflow controls.

  1. Article

    Published article: Why “Almost Human, But Not Quite” Feels Wrong

    Why clowns, AI-generated faces, and fluent AI text can feel uncanny when human-like cues conflict with emotion, realism, or factual coherence.
  2. Article

    Published article: LLM-Led vs Orchestrator-Led Tool Execution

    Comparison of LLM-led and orchestrator-led tool execution in LLM systems, including control-plane placement, reliability, observability, latency, cost governance, and security.
  3. Article

    Published article: LLM Sycophancy: Evidence, Training Pathways, and Evaluation

    Learn what LLM sycophancy means, how user beliefs and confidence can shift answers, and how to test agreement pressure and correction selectivity.
  4. Article

    Published article: Prompt Engineering Guide for Daily Work

    Learn five AI prompt failure modes and the evidence rules, coverage checks, tool policies, and acceptance tests that make professional outputs reliable.