Data analyst | Blockworks
Joined August 2021
- Tweets2K
- Following443
- Followers620
- Likes8.7K
My takeaways:
1) The Orchestrator-subagent swarm is overrated!
2) A curated set of custom skills, tools and Memory .MD outperforms both Vanilla Claude and Swarms.
A good Subagent I'm still using :
A tool call that invokes a Sonnet subagent to read Anthropic documentation. It spares my main agents' context window when I ask it "Can Claude do X?"
A good skill file I'm maintaining:
A /dashboard-preferences skill that tells my agent how to make my data dashboards look clean.
When building your own skill files and memory files, you can prevent context bloat by storing them in a project folder. You can also add disable-model-invocation if you don't want its description to always appear in your agents context window.
One problem I'm still facing is maintaining up to date Readme files and other project documentation as new decisions get made.
claude.com/resources/article…
Anthropic is now recommending that companies REMOVE unnecessary subagents from their AI systems.
This is a pretty significant shift in agent architecture, with their team finding that a single agent equipped with skills can frequently outperform complex multi-agent systems, and one customer seeing roughly 2x faster execution.
The emerging rules of thumb:
→ Skills: Teach the agent HOW to do something, including procedures, expertise, and instructions.
→ Tools / MCP: Give the agent the ability to DO things through APIs, actions, and live data.
→ Subagents: Use for parallel work, isolated context, or tasks a single agent struggles with.
The interesting part is progressive disclosure. Instead of spawning specialized agents or stuffing everything into a giant prompt, one agent discovers and loads skills and tools as needed.
Their advice is to fold existing subagents into the main loop, turn their instructions into skills, expose their actions as tools, and run evals to see if performance holds up.
As models have gotten smarter, improving agent architecture is becoming less about what you can add and more about what you can delete.
claude.com/resources/article…
Each Ripple $RLUSD is backed by $1.07 !
Track reserve info in our stablecoins sector dashboard below:
blockworks.com/analytics/sta…
Mantle is refreshing its branding. They are doing cool stuff on the RWA side. All the xStocks universe is available on Mantle. They also have some pre-IPO offerings. 👀
Thank you @TheCompoundNews for showcasing my charts on Animal Spirits.
I liked the episode. Michael was passionate! 🔥
Say bubble one more time. We dare you.
On Episode 474 of Animal Spirits, @awealthofcs and @michaelbatnick break down the tech stock bloodbath, why the Mag 7 isn't carrying the market anymore, South Korea's wild investor behavior, whether Nike lost its moat, why boomers aren't downsizing, stocks vs. housing, Netflix's slide, the return of movies, and more.
Watch/listen now👇🔥
Welcome Shards #12 and #13 !👶
NEAR's dynamic sharding automatically scaled the network up based on demand.
New shard #12 is already consuming the most gas.
The 2.13 upgrade also brings dynamic resharding to mainnet.
When a shard fills with state, it splits automatically. No vote, no validator coordination, no protocol upgrade.
NEAR's capacity is now a property of the network, not a governance decision.
near.org/blog/introducing-dy…
I also tested whether prior market returns were related to how often “bubble” came up.
The relationship was weak, although Animal Spirits showed a modest positive link with prior-quarter returns.
Translation: stronger markets invite more bubble talk, but the signal is weak.
I put all the charts, episode-level data, and correlations into a pair of vibe-coded dashboards:
incredible-semolina-9f6893.n…
Robinhood tokenized stocks are generating a lot of buzz. Rialto is a new venue for trading these stocks.
I wrote this Dashboard Primer on the BNB Smart Chain. Please check it out and let me know what you think. 😀
xitter.cf/blockworksres/status/2…
