Guides

1. EU AI Act: Duties for Agent Deployers

   We deploy an agent to EU users. Which published AI Act obligations attach to us rather than to the model provider?

   Find the Commission’s role distinctions, obligation sets and application dates, then record which questions about your agent need counsel.

   Read the guide →
   Scope · 5 topics

   • Role identification for agent systems and underlying models
   • Published deployer versus provider obligation sets
   • Application dates and future high-risk requirements
   • Hypothetical third-party agent deployment
   • Questions requiring counsel and weekly maintenance

2. Agent Data Retention: Review the Actual Terms

   If I point an agent at our code and customer tickets, what do the vendors' published terms say happens to that data by default?

   Review training, retention, human access and data transfers for the account receiving your code and tickets, with a contract-specific approval record.

   Read the guide →
   Scope · 6 topics

   • Separate training use, abuse monitoring, application-state retention, human access and onward transfers
   • Review OpenAI API retention periods, deletion triggers and control exceptions
   • Read Anthropic commercial restrictions without inventing a retention schedule
   • Map Copilot CLI content and telemetry destinations without assuming plan defaults
   • Resolve a hypothetical enterprise-to-personal-account coverage assumption
   • Maintain a contract-specific approval record with weekly source review

3. Build an Agent Regression Set Before Adoption

   What should I measure on my own work before and after adopting an agent, so a later change is provable?

   Freeze real tasks, define who judges success, and compare outcomes, escalations and failures before and after an agent enters your workflow.

   Read the guide →
   Scope · 8 topics

   • Building a reader-owned evaluation set before agent adoption
   • Recording known-good outcomes, judgment criteria and reviewers
   • Comparing resolution, escalation, human effort, time and cost
   • Retaining task-level failures and run evidence
   • Separating capability failures, permission failures and ambiguous tasks
   • Maintaining frozen baselines alongside evolving task collections
   • Interpreting hypothetical support results without claiming productivity gains
   • Reviewing framework references monthly

4. Agent Identity: API Keys and Delegated Tokens

   How does an agent authenticate to my systems, and what breaks when it holds a long-lived key instead of a delegated token?

   Separate user login, server access and downstream credentials, then assess scope, lifetime, revocation and what an incident reviewer could reconstruct.

   Read the guide →
   Scope · 8 topics

   • User login, client-to-server identity and downstream authorization
   • Shared API keys versus per-user delegated credentials
   • Audience validation, token separation and confused-deputy risks
   • Credential storage, access-token expiry and renewable grants
   • Revocation boundaries and integration disruption
   • Agent attribution versus human-requester attribution
   • Six-month incident-review evidence
   • Documentation limits and monthly maintenance

5. Calculate Agent Cost per Completed Task

   How do I turn per-token and per-seat agent pricing into a defensible cost per completed task?

   Build a defensible agent budget from measured tokens, retries, successful outcomes and seat costs, with a synthetic worksheet you can adapt.

   Read the guide →
   Scope · 8 topics

   • Defining completion and distinguishing tasks, runs and requests
   • Instrumenting tokens, cache usage, retries and tool-call round trips
   • Mapping published rates and tool-specific billing quantities without double-counting
   • Calculating cost per attempted run and successful task with visible completion rates
   • Synthetic arithmetic showing why half-price runs can cost more per completed task
   • Allocating seat costs and subscription-included usage over a matching period
   • Separating rate limits and wall-clock costs from token prices
   • Building auditable budget scenarios and maintaining the canonical method monthly

6. How to Audit a Coding-Agent Benchmark Claim

   A vendor advertises a coding-agent benchmark score. How do I check whether it means anything for my repository?

   Check the task set, harness, scaffold, model and attempt budget behind a coding-agent score before deciding what to evaluate on your repository.

   Read the guide →
   Scope · 7 topics

   • Reconstructing a vendor claim and marking missing methodology
   • Distinguishing benchmark, variant, harness, scaffold and model
   • Recording model versions, evaluation dates and execution budgets
   • Separating attempts per task, turns, batch selection and aggregation
   • Exposing a hypothetical full-set versus filtered-subset comparison
   • Assessing repository relevance without forecasting productivity
   • Documentation limitations and monthly canonical maintenance

7. Agent Autonomy: Approvals, Sandboxes and Stop Controls

   When a vendor calls an agent "autonomous", which approval and sandbox controls am I actually being given?

   Read autonomy claims critically: separate approval gates, allowlists, execution boundaries, output review and stopping controls for the exact product.

   Read the guide →
   Scope · 9 topics

   • Autonomy in deciding versus autonomy in acting
   • Human approval, classifier review and reusable permissions
   • Filesystem, network and credential boundaries
   • Input, output, tool and logging-only guardrails
   • Product-surface scope and undocumented defaults
   • Output destinations and review gates
   • Mid-flight stopping and execution limits
   • Hypothetical proposal-only and isolated pull-request agents
   • Five-part buyer control record

8. Audit an MCP Server Before You Install It

   Before I install an MCP server, how do I work out what it can actually reach and who can steer it?

   Map declared tools, credential permissions, host approvals and untrusted inputs, then record an approve, restrict or refuse decision.

   Read the guide →
   Scope · 9 topics

   • Pre-install inventory of tool names, complete schemas and parameters
   • Declared behavior versus handler reach and credential entitlement
   • Local execution and remote discovery boundaries
   • Host filtering, deferred loading and confirmation policy
   • Untrusted-content influence and who controls each boundary
   • Hypothetical ticket-server blast-radius audit and narrower redesign
   • Confused-deputy consent and token-passthrough review
   • Approve, restrict or refuse decision record
   • Weekly canonical-guide maintenance and evidence limitations

9. Does Claude Code Support AGENTS.md?

   Compatibility · 6 min read

   No. Claude Code reads CLAUDE.md natively. Here is the official symlink and import workaround, and which coding agents in our index already read AGENTS.md on their own.

   Read the guide →
   Scope · 5 topics

   • Claude Code
   • AGENTS.md
   • CLAUDE.md
   • Coding agents
   • Compatibility

10. Framework, Platform, or Ready-Made AI Agent?

    Choosing · 15 min read

    The decision that comes before any tool comparison: what you are actually buying at three levels of the stack, and the tradeoff at each one.

    Read the guide →
    Scope · 4 topics

    • Frameworks
    • Agent platforms
    • Decision guide
    • Build vs buy

11. n8n vs Lindy vs Dify: 16 AI Agent Platforms Compared

    Choosing · 20 min read

    All 16 agent platforms in our index, split on what you are really choosing: open-source developer tool, no-code builder, managed enterprise, or personal assistant.

    Read the guide →
    Scope · 18 topics

    • Agent platforms
    • Decision guide
    • n8n
    • Lindy
    • Relevance AI
    • Dify
    • Botpress
    • Sierra
    • Salesforce Agentforce
    • Moveworks
    • Glean
    • Botsify
    • Vecbase
    • OpenClaw
    • MindStudio
    • Sim
    • Vellum
    • ElevenLabs

12. How to Choose an AI Coding Agent: 25 Tools, 4 Jobs

    Choosing · 20 min read

    Coding agents is our biggest category at 25 tools, and they split into four genuinely different jobs, so “which is best” is the wrong question.

    Read the guide →
    Scope · 10 topics

    • Coding agents
    • Decision guide
    • Cursor
    • Claude Code
    • Devin
    • OpenCode
    • CodeRabbit
    • Zed
    • Goose
    • Kilo Code

13. AI Support Agents: Decagon vs Sierra vs Intercom Fin

    Choosing · 14 min read

    Every customer-support agent we index, compared: six genuinely different ways to buy one, and the right pick is whichever you actually want.

    Read the guide →
    Scope · 10 topics

    • Support agents
    • Decision guide
    • Decagon
    • Sierra
    • Intercom Fin
    • Zendesk AI Agents
    • Crescendo
    • Cresta
    • Ada
    • Botsify

14. How to Choose an AI Research Agent: 11 Compared

    Choosing · 17 min read

    The 11 research agents in this guide look like rivals, but each is built for a different job: web answers, literature, legal, raw web data, or your own data.

    Read the guide →
    Scope · 13 topics

    • Research agents
    • Decision guide
    • Perplexity
    • Manus
    • Genspark
    • GPT Researcher
    • Webhound
    • Elicit
    • Undermind
    • GC AI
    • Microsoft Fabric Data Agent
    • Julius AI
    • TinyFish

15. 11x vs Artisan vs Lead Scorer: 8 AI Sales Agents Compared

    Choosing · 16 min read

    All 8 sales-and-marketing agents in our index: eight tools solving eight different jobs at eight different budgets, not eight competing chatbots.

    Read the guide →
    Scope · 10 topics

    • Sales agents
    • Decision guide
    • 11x
    • Artisan
    • Qualified
    • Rox
    • Unify
    • Warmly
    • Amplemarket
    • Lead Scorer

16. AI Voice Agents: Vapi vs Retell AI vs ElevenLabs

    Choosing · 19 min read

    Every voice agent we index, compared by the layer of the stack you own: raw model, developer platform, no-code builder, or managed incumbent.

    Read the guide →
    Scope · 13 topics

    • Voice agents
    • Decision guide
    • OpenAI Realtime API
    • Vapi
    • Retell AI
    • Bland AI
    • ElevenLabs Conversational AI
    • Synthflow AI
    • LiveKit Agents
    • PolyAI
    • Deepgram Voice Agent API
    • Cartesia
    • VoiceAgent

17. LangChain vs CrewAI vs 12 More Agent Frameworks

    Choosing · 17 min read

    All 14 agent frameworks in our index side by side: what each is actually for, live GitHub stars, and how to pick one for what you’re building.

    Read the guide →
    Scope · 14 topics

    • Frameworks
    • Decision guide
    • LangChain
    • LangGraph
    • CrewAI
    • LlamaIndex
    • AutoGen
    • Microsoft Agent Framework
    • Mastra
    • Letta
    • Google ADK
    • Claude Agent SDK
    • Agno
    • AgentsKit

18. AI Agent Pricing in 2026: Every Tool in Our Index

    Pricing · 13 min read

    The pricing model, free tier and API access for every AI agent we index. Freemium leads, but usage-based billing is nowhere near an edge case.

    Read the guide →
    Scope · 3 topics

    • Pricing
    • Market data
    • Directory census

New guides appear here when they are published. Follow the RSS feed.