Getting stuff done
$ ls -la ~/posts/
drwxr-xr-x
DATE
NAME
-rw-r--r--
Feb 05
Not all context is equal: hierarchical memory for AI agents
-rw-r--r--
Feb 03
Cache the reasoning, not the answer
-rw-r--r--
Jan 30
Local, long term memory for OpenClaw agents
-rw-r--r--
Jan 27
From facts to insights: how observations work in Hindsight
-rw-r--r--
Jan 25
What learning actually means for AI agents
-rw-r--r--
Jan 16
File-based agent memory: great demo, good luck in prod
-rw-r--r--
Dec 24
Background operations: what happens after retain()
-rw-r--r--
Dec 23
Temporal reasoning: "when it happened" vs "when you learned it"
-rw-r--r--
Dec 22
Document upserting: keeping evolving conversations fresh
-rw-r--r--
Dec 21
TEI for production: embeddings and cross-encoder reranking
-rw-r--r--
Dec 16
Beyond vector search: how TEMPR combines 4 retrieval strategies
-rw-r--r--
Dec 15
Rich fact extraction: preserving narrative, not just statements
-rw-r--r--
Dec 14
Retain, recall, reflect: the three operations of agent memory
-rw-r--r--
Dec 13
Opinions with confidence scores: how agents form beliefs
-rw-r--r--
Dec 12
Memory types in Hindsight: world, experience, opinion, observation
-rw-r--r--
Dec 12
Drop SQLite: zero-dependency quick starts with pg0
-rw-r--r--
Dec 10
Token budgets vs top-k: a better way to fill context windows
-rw-r--r--
Dec 08
Hindsight vs traditional RAG: what you actually get
-rw-r--r--
Nov 26
pg0: zero-dependency PostgreSQL for development
-rw-r--r--
Nov 16
The reasoning agent: a different architecture for AI systems (part 1)
-rw-r--r--
Nov 10
LongMemEval: debugging a 300MB JSON file dataset
-rw-r--r--
Nov 05
Code comments: humans vs agentic code$ wc -l ~/posts/*
22 posts total
$ ls ~/tags/
$