Explore a topic
Knowledge management and research
Find information, check its sources and keep useful notes easy to find and update.
Start with a practical guide
Start with a collection
Compare all collections on this topic
Keep knowledge so you can find, trust and reuse it later
Design a lightweight personal knowledge system around future use: admission rules, realistic retrieval cues, preserved context and provenance, canonical notes, incremental deduplication, explicit stale states, decision bundles and maintenance triggered by real reuse.
Search for evidence as a reproducible process, not a lucky query
Find stronger evidence by decomposing concepts, testing retrieval, chaining citations, documenting searches and using AI as a search assistant without letting generated summaries become the evidence.
Trace where information came from before deciding whether to trust it
Separate provenance, authenticity and factual support so repeated or AI-mediated information can be verified from accountable origins rather than from appearance or citation volume.
Give AI the right context, memory and reusable skills
Use modern context engineering, durable memory and evaluated skills to make repeated AI work more reliable without stuffing every fact and instruction into every conversation.
Turn mistakes and outcomes into changes you can verify
Add a compact learning loop for debriefing, root-cause exploration, prospective failure analysis and stronger corrective actions without collapsing every failure into individual blame.
Ask questions that produce usable evidence instead of polished generalities
Turn an unclear information need into neutral questions, concrete examples, timelines, artifacts and explicit unknowns that can change a decision.
63 ideas
Treat context as a budget, not a transcript
A bigger context window is not permission to turn the whole project into one prompt.
Read the idea PrincipleSeparate session state from long-term memory
What the agent needs for the next hour is not the same thing it should remember next month.
Read the idea ProtocolGive agent memory a write-manage-read policy
Memory is a loop, not a folder.
Read the idea ProtocolEvaluate what memory retrieves, not just what it stores
A perfect archive can still produce a bad working memory.
Read the idea ProtocolKeep raw evidence behind AI memory summaries
A summary should be a door to the evidence, not the place where the evidence disappears.
Read the idea PrincipleUse progressive disclosure for agent instructions
Do not make every future task pay the context cost of every past instruction.
Read the idea HeuristicTurn a recurring AI workflow into a skill only when the trigger is clear
A reusable skill should remove repeated thinking, not create a new library you now have to search.
Read the idea ProtocolAblate a skill before you trust it
If removing the skill changes nothing, the skill may be documentation for humans rather than leverage for the model.
Read the idea ProtocolCompare expected and actual before explaining the gap
Start the review with two columns, not a culprit.
Read the idea ChecklistBring objective traces into the debrief
Memory is useful input; it should not be the only log.
Read the idea ProtocolAsk what helped, what hindered and what changes next
A useful review connects observation to a specific next change.
Read the idea TemplateGive the lesson a verifiable finish
A corrective action needs a finish line that can be inspected.
Read the idea ProtocolFollow the why-chain until the cause becomes changeable
Do not stop at a label that merely renames the event.
Read the idea ProtocolFan possible causes out before narrowing them
Widen the causal search before selecting the story.
Read the idea ProtocolRun failure-mode analysis before the consequential change
Ask how the process can fail before production teaches you.
Read the idea HeuristicPrefer a system-strengthening action over another reminder
If the same slip can recur unchanged, the action probably did not reach the system.
Read the idea ConceptSave information for a future job, not because it is interesting
Keeping has a cost, so give it a future need.
Read the idea ProtocolWrite one retrieval cue in the title
The future query may not resemble the source's headline.
Read the idea PrincipleStore why the item mattered when you saved it
The missing information is often your old context, not the page.
Read the idea ProtocolKeep source, locator and access date beside a material claim
A claim ages better when verification is one step away.
Read the idea PrincipleSeparate what the source says from what you infer
Your future self deserves to know which sentences came from evidence and which came from you.
Read the idea ProtocolGive changing knowledge an explicit state
A knowledge base needs state, not only storage.
Read the idea ProtocolSearch before creating a second canonical note
Duplicate knowledge creates maintenance forks.
Read the idea ProtocolUse aliases for names that change but the knowledge object does not
Names drift faster than the underlying need.
Read the ideaExplore another topic
Critical thinking and evidence
Test claims, separate evidence from interpretation and ask questions that can change what you believe.
70 ideasLearning and skill building
Practice a skill, check what you remember and use feedback for the next attempt.
90 ideasWorking with AI
Give AI a clear task and context. Check its work and set limits on what it can do.
133 ideas