The fenbs blogAll articles
Page 24 of 48
- Roles and permissions for humans and AI agents, on one boardOne list of roles, held by people and AI assistants alike. What each permission allows, how the ceiling works, and why a refusal is a full sentence. · 7 min read
- An audit trail for AI agents: knowing who changed whatIf an AI agent can change your board, every change needs its name on it, and that record has to survive the day you take its access away. · 6 min read
- Project management for construction site teams: a board a foreman will actually useA small site team needs one shared list of what to build, what changed and what needs putting right, and a way for the crew to mark it done. · 7 min read
- How to connect Claude to your project board with MCPOne command in Claude Code, or one connector in Claude on the web. Then a first question, a first task, and how to take the key back. · 6 min read
- Features, enhancements and bugs: a simple way to organize workAlmost all work is adding something, improving something or fixing something. That is the method, and it is where the name fenbs comes from. · 6 min read
- AI Context Document: A Template for Clear Project BriefsAn AI context document gives an assistant the goal, rules, sources, current decisions, and next step. Use this template to brief a new session without retelling the whole project. · 5 min read
- Email Writing Prompts for AI: Clear Replies and Follow-UpCopy these prompts for updates, requests, follow-ups, and difficult replies. Give the assistant the facts and limits it needs, then review the message before sending. · 5 min read
- Android AI Development: A Reviewable App WorkflowGive an AI coding assistant the Android project context it needs, ask for small changes, and verify behavior on devices. Keep requirements consistent across sessions. · 5 min read
- AI Prompt Engineer: What the Role InvolvesAn AI prompt engineer turns a vague request into instructions a model can follow, then checks whether the result works. Here is the practical work, the skills it needs, and a useful first project. · 5 min read
- ChatGPT Project-Only Memory Not Working: Test the ScopeCheck the actual project setting, required memory controls, and source of an unexpected answer. A controlled test helps separate a boundary problem from a missed detail. · 5 min read
- Claude Memory Compaction: What Survives a Long ChatCompaction shortens the working conversation; saved memory is a separate mechanism. Learn what to check after a long chat is summarized and how to preserve the next step before restarting. · 5 min read
- Company Brain With Supermemory: A Practical EvaluationEvaluate Supermemory for company knowledge retrieval, distinguish the memory engine from its former hosted Company Brain, and keep approved work clear. · 5 min read
- Does Claude Remember Deleted Chats? Check What RemainsDeleting a Claude chat and deleting information remembered from it are separate checks. Review memory, project material, and any work records you saved elsewhere. · 5 min read
- How to Give Claude a Previous ConversationGive Claude a checked transcript or summary with your current goal, approved decisions, and unresolved work clearly separated. Verify its understanding before asking it to continue. · 5 min read
- ChatGPT Memory Import: Supported Paths and a Manual FallbackCheck which kind of memory you are moving and which destination supports it. A reviewed context summary is useful, but it is not a restored memory database. · 5 min read
- Notion Wiki: Build Knowledge People and AI Can UseBuild a Notion wiki around clear questions, named owners, and approved answers. Keep reference pages useful while separating them from active decisions and work handoffs. · 6 min read
- Second Brain Extensions: What to Save and How to ChooseCompare browser extensions for capturing web knowledge, then check retrieval, privacy, and portability before you trust them with useful work context. · 8 min read
- Second Brain Websites: Choose a Home for Your AI WorkChoose a web-based second brain by the material you save, the people who need access, and the work you need to continue. Compare Notion, Craft, Capacities, and fenbs by their best fit. · 8 min read
- AI-Driven Automation: Keep Work Moving With Clear HandoffsAI-driven automation can interpret routine work and choose a next step. Make each transition visible so a person or another assistant can safely continue it. · 5 min read
- AI Mobile App Development: From Requirements to Device TestsUse AI to draft and revise a mobile app in small, reviewable changes. Keep requirements, device checks, and release decisions attached to the work. · 5 min read
- How to Ask AI Questions and Check the AnswersAsk a specific question, supply the evidence that matters, and check the claims before using the answer. Save verified references and open questions so the next conversation can build on them. · 6 min read
- ChatGPT for Compliance: Research, Checks, and Human ReviewUse ChatGPT to prepare source-led compliance drafts, then verify the evidence and route interpretations to the right person. Keep approval separate from generation. · 5 min read
- Claude Haiku 4.5 Context Window: What Fits in a Session?Haiku 4.5 has separate specifications for context capacity and maximum output. Match them to your access route, choose relevant inputs, and preserve the project record when you start again. · 5 min read
- Claude System Prompt: What It Is and What You Can SetSeparate Anthropic’s published system instructions from the prompts and settings you control. Learn where to put stable work rules and how to check their effect. · 6 min read
- Copilot Memory in VS Code: Inspect Tools, Files, and ScopeInspect the saved notes used by your VS Code agent, test what survives a new conversation, and keep local memory separate from shared project instructions. · 5 min read
- Does Copilot Have Memory Like ChatGPT? Check Your AccountMicrosoft Copilot offers memory and personalization, but the account and app determine which controls apply. Compare its retained context with ChatGPT without confusing personal memory, work access, and chat history. · 7 min read
- Gemini Prompts for Work and Study You Can ReuseUse these adaptable Gemini prompts to summarize sources, practice a subject, plan work, and review a draft. Each includes a clear boundary and a way to check the result. · 5 min read
- How to Turn Off Claude MemoryPause memory for your account or turn it off before starting one chat. Understand the difference between pausing, deleting memories and disabling past-chat search. · 5 min read
- Personalized Learning With AI: Build a Study WorkflowAdapt explanations, practice, and feedback to what you need to learn. Keep evidence of progress and useful lessons between study sessions. · 5 min read
- Second Brain GitHub Projects: Choose an Open Source SetupChoose an open source second brain by the work you need to recover, the notes you want to own, and the maintenance you can actually do. · 8 min read
- Slack AI Bots: Turn Team Questions Into Follow-Up WorkChoose the right kind of AI help in Slack, check its answers, and turn a useful conversation into decisions and tasks your team can continue. · 5 min read
- AI Background Knowledge: What It Knows and What to SupplyAn AI may understand your field without knowing your project. Separate general knowledge from current sources, private facts, and the decisions only your team can supply. · 6 min read
- AI for Report Writing: From Evidence to a Checked DraftUse AI to organize a work report around verified evidence. Keep findings separate from assumptions, check every important claim, and leave a clear next drafting task. · 5 min read
- AI Governance Roadmap: From Inventory to Ongoing ReviewBuild AI governance in phases with a clear owner, a useful deliverable, and an exit check for each stage. Keep implementation separate from a policy launch. · 6 min read
- AI Introduction Writers: Draft an Opening That FitsGive an AI introduction writer your audience, purpose, and actual source material. Compare a few focused openings, then choose one that accurately introduces what follows. · 5 min read
- AI Organizer: Choose a Home for Notes, Tasks, and ContextChoose an AI organizer by the work you need to recover: notes, actions, or the current project story. Compare practical options and test one workflow before moving everything. · 8 min read
- AI Prompts for Lawyers: Drafting and Source ChecksUse legal work prompts with a defined scope, supplied authority, and an explicit verification record. Copy templates for issue maps, document review, drafting, and checking tasks. · 5 min read
- AI Web Development: From Clear Requirements to Tested ChangesUse a coding assistant to build and revise a website in small, reviewable steps. Keep the requirements, acceptance checks, and unfinished work clear across sessions. · 5 min read
- AI Writers With Citations: How to Check Their SourcesA citation is a route to evidence, not a guarantee that a claim is correct. Compare source-supported writing workflows and learn a repeatable way to check the references before using a draft. · 8 min read
- ChatGPT vs Claude for Government Work: What to CompareStart with the approved use, deployment and procurement requirements. Compare ChatGPT and Claude on a permitted pilot, with a person reviewing the evidence before any result affects the public. · 7 min read
- .claude.json at the Project Level: Use the Right FileProject-specific entries in .claude.json are local state, not a shared project handbook. Learn where settings, instructions, and changing work context belong. · 5 min read
- Claude Memory Not Available? What to CheckFind out whether memory is missing because of your account, organization, chat mode or app surface. Keep a deliberate project handoff while you investigate. · 5 min read
- Copilot Pages vs ChatGPT Canvas: Editing and Shared WorkCopilot Pages turns drafts into shared Microsoft 365 pages. Compare that workflow with the original ChatGPT Canvas and the editing options ChatGPT offers now, including what happens to sources and approvals. · 7 min read
- Notion Knowledge Base AI Assistant: Organize and AskMake Notion knowledge pages easier to question with AI. Organize the sources, limit the question, inspect citations, and keep approved decisions available beyond the answer. · 5 min read
- AI Rules: Set Clear Boundaries for Your AssistantsUseful AI rules tell assistants what they may do, what they must check, and when a person must decide. Separate those working instructions from company policy and legal requirements. · 5 min read
- Secure Knowledge Management: Set Boundaries for AI WorkDecide what knowledge an assistant may use before connecting it. Classify information, check access and vendor controls, and keep approved rules separate from sensitive source material. · 5 min read
- What Is an AI Governance Layer? Rules, Controls, and EvidenceAn AI governance layer connects the rules people approve with the controls that apply them and the evidence that shows what happened. · 5 min read
- AI Memory Decay: When Remembered Information Becomes StaleAI memory decay can mean information disappears, becomes harder to retrieve, or remains available after it stops being true. Each problem needs a different response. · 5 min read