Ask HN: Do you also "hoard" notes/links but struggle to turn them into actions?

That’s awesome — building it is the easy part, making it a habit is the real test.

What made you actually keep using it so far: low friction, reliable time saved, or tight integration into something you already check daily?

I wrote up the core loop I’m exploring in my HN profile/bio — would love to hear how your agent differs (especially: where it pulls context from, how it outputs actions, and how it avoids becoming noisy).

I’m with you on “systems don’t guarantee outcomes” — the only reliable commonality is shipping (or at least producing something real) over time.

One thing I did take away from your posts is the distinction between the medium you think in (pen/paper) and an “authoritative knowledge system” that’s the source of truth for your life/work. That framing resonates: capture can be messy and human, but “what is true / what is next” needs a home you trust.

That’s pretty close to what I’m poking at with Concerns: not “optimize PKM”, but “reduce the gap between what you already know and what you actually do next”. The constraint I’m leaning on is forcing a small number of active goals/projects, so the system can’t pretend everything matters — it has to help you commit.

Curious: in your setup, what’s the smallest digital thing you’d accept as the authoritative source of truth (calendar? a single task list? a project index)? And what’s the trigger that turns a note into an action for you?

This is a great example of making the system trustworthy: Inbox → scripted processing → git diff review → reset if needed. The diff-as-trust-layer feels like the missing piece for a lot of “AI organizes my notes” ideas.

Also +1 on “limited rules” being more important than fancy models — once the structure is predictable, Claude becomes a genuinely useful retrieval and research assistant.

Out of curiosity, where does the loop close for you today? After “deep research” produces a conclusion, do you write back into a MOC / decision log / project note, or does it mostly live in chat + commits? I’m exploring a similar loop (context → retrieval → suggestion → human review → write-back) with the same constraint: keep it auditable and reversible (diff-friendly). Details are in my HN profile/bio if you want to compare approaches.

I hoard information that I find interesting but cannot act on right now, mostly because engaging with it immediately would derail me. The problem is that I almost never come back to what I have saved.

A simple example is YouTube. I save videos to watch later because I am not in the right headspace at the time. Then I avoid them completely. I think part of the resistance is that I know watching them properly will demand attention and probably lead to follow-up work, and I am rarely in a mode where I want that interruption.

I have thought about the whole “second brain” idea, but I worry it would just become a dumping ground. Nothing would really resurface when it actually matters. I would mostly be relying on myself to remember that I once made a note about something when I happen to be working on a related problem.

Lately I have been thinking about the idea of a passive, radio style feed that summarises the information I have collected and plays it back to me, so I can at least consume each item once.

You see those TV shows about people who hoard. They cannot throw things away because they might be important one day. This feels uncomfortably similar.

Maybe the real problem is not how we store information. Maybe it is that we aren't filtering hard enough on what is actually worth keeping in the first place.

I’ve recently started using a similar process to what you’re describing. It basically involves writing/speaking to AI in an unstructured manner, and having it extract entries for my knowledge base based on a certain schema. I built a tool to validate the entries to the schema, giving the AI a feedback loop so it can automatically fix/improve the entries. It also supports linking. I use agentic search in both entering and finding knowledge. Links, categories and metadata help the AI tremendously in finding relevant snippets. I go into more detail in this blog post [0].

[0] https://thalo.rejot.dev/blog/plain-text-knowledge-management