RecommendationNikolay Petrov recommends evaluating a skill by running two arms — one with no skill loaded and one with the skill loaded (providing additional information to context) — and comparing them on multiple metrics: how many evals are solved, token usage, and time to complete tasks. Depending on the case, the comparison can run with no tools to check the model's internal understanding, or with tools like MCP access for a more practical setup.
AI Engineer · Your SKILL.md Is 70% Stuff the Model Already Knows - Nikolay Petrov, Intel
RecommendationNikolay Petrov recommends a deterministic process based on the idea of atoms from fact score: instead of relying on luck, split the skill and its instructions into atoms — roughly 25-30 per skill — and for each atom create a pair of question and expected answer, either general propositions or contracts with definitive expectations on the action. He also advises ensuring the questions do not leak their answers.
AI Engineer · Your SKILL.md Is 70% Stuff the Model Already Knows - Nikolay Petrov, Intel
Recommendation · MoneySparsh Jain recommends building an experiential memory layer into AI agents so they accumulate judgment from mistakes and course corrections. Divide agent memory into short-term (working/active session) and long-term (fact-based and experiential) memory, and structure the experiential layer into three categories: case-based, strategy-based, and skill-based. Store detailed cases and detailed learnings in the case-based category, store situation-specific strategies that worked in the strategy-based category, and abstract generalizable learnings into reusable skills the agent can apply repeatedly.
AI Engineer · Your AI Agent Keeps Making the Same Mistakes - Sparsh Jain, CentralAgent AI
Recommendation · MoneyCohen advises that almost for sure you started with prices too low — the microeconomics demand curve applies, but most companies are on the wrong side of it, not at the price ceiling. Raising prices works because price determines which market you're in: hobbyists drop out, but the middle market that wouldn't consider a $10 product starts considering a $50-$500 one — a bigger market with more budget and far lower churn (1.5%/month vs 9%/month). A V1 product can justify lower pricing because it does less, but that excuse expires; two years later it's not a V1 and you need to charge for what it is.
hedged
Startups For the Rest of Us · Episode 857 | Finding Hidden Multipliers in Your Business (with Jason Cohen)
Recommendation · MoneyCohen advises positioning your product around creating more of the value the customer already wants, rather than around saving them time, money, or effort — you can charge up to 10x more with a value-creation pitch. A 'save money' pitch caps your price at roughly 10-20% of the savings, because charging more than a small fraction of what you save makes the deal not worth it. With a value-creation pitch, customers pay for the whole value created: if a customer already spends $40K on leads and you double their leads, you can charge up to the full extra $40K because their spending behavior already proves that's what doubling leads is worth to them.
Startups For the Rest of Us · Episode 857 | Finding Hidden Multipliers in Your Business (with Jason Cohen)
Recommendation · MoneyCohen advises identifying 'keystones' — the one or two things a customer likes so much they drag themselves through the trade-offs they don't like — as the main reason someone buys, rather than a massive rubric. Identify 'inciting events' — the trigger that puts a buyer in-market — because nobody randomly wakes up and decides to replace their CRM or move hosts; there's always a reason why now. Identify 'deal breakers' — reasons a prospect who loves everything still can't buy (e.g., a big company won't buy a small product that isn't SOC 2 compliant because governance policy mandates it). Describe your ideal customer by behaviors and attitudes, not firmographics/demographics — company size is 'almost never the right way'; e.g., Kanban buyers are defined by wanting to move quickly, being lean
Startups For the Rest of Us · Episode 857 | Finding Hidden Multipliers in Your Business (with Jason Cohen)
Recommendation · MindHuberman ranks morning bright light, resistance training, and maximizing sleep quality as the top three things to do for a healthier life.
The a16z Show · Andrew Huberman’s Three Rules for Better Health and Focus
Recommendation · MindHuberman recommends protecting your two daily focus windows (he cites roughly 9:30 a.m.–12:30 p.m. and ~4:30–7 p.m.) as if they were the most important times of the day, with access limited to family and coworkers with emergencies.
The a16z Show · Andrew Huberman’s Three Rules for Better Health and Focus
Recommendation · MoneyThe process involves taking written material, often a skill MD file with scripts, references, and assets, and moving it into an actual graph structure. Giving the agent access to a graph structure with nodes representing skill decisions, routing points, and scripts to be executed, and forcing the agent to run that graph, results in increases in task reward and performance. Compiling prose down to deterministic scripts is part of the process that leads to increased task reward and performance.
AI Engineer · Make Your AI Agents Predictable With Graph Engineering - Zachary Blumenfeld, Neo4j
Recommendation · MoneyYou can take a context graph and distill a skill from it, keeping it connected in the same graph to all past decision traces and data. Distilling a skill from a context graph gives you performance benefits plus provenance back from every step, versioning, tracking, and governance over time where you can assess drift. You can understand where certain tool calls are coming from and their history when using graph-based skills.
AI Engineer · Make Your AI Agents Predictable With Graph Engineering - Zachary Blumenfeld, Neo4j
Method · MoneyZachary Blumenfeld describes compiling the information from successful agent runs into a graph YAML shape inside a skill, and giving the agent a CLI to run that programmatic skill compiled from documentation or the memory graph, rather than building the loop inside the agent.
Method, not a verified protocol.
AI Engineer · Make Your AI Agents Predictable With Graph Engineering - Zachary Blumenfeld, Neo4j
Recommendation · MoneyKatie Givens recommends pulling six to eight quarters of baseline data so you can see trends at the organization rather than judging a single snapshot, and analyzing sales metrics on a 6-8 quarter trend line because that's where the most informed decisions come from. She uses the trend to establish a baseline and see where the organization has dropped off — there is no single 'right' number for metrics like sales-cycle length.
30 Minutes to President's Club | No-Nonsense Sales · #622: Sales Enablement Playbook: How To Win Your First 90 Days
RecommendationAlex Cooper recommends that instead of asking a pregnant woman if she's 'so excited,' you ask an open-ended question about how she's feeling, leaving room for her to say she is grateful and also scared out of her mind.
Call Her Daddy · The Unknowns of Motherhood
RecommendationAlex Cooper recommends that instead of asking a pregnant woman 'are you so excited?', you ask an open-ended question like 'how are you feeling?' so she can answer honestly — whether excited or struggling. An open-ended question allows a pregnant woman having a good day to share joy (e.g., 'I just went and got our nursery all set up') and allows a struggling pregnant woman to share honestly (e.g., 'I have been dry heaving over the toilet between meetings'). She also recommends giving pregnant women permission to say both: 'I am so incredibly grateful. And I'm also really scared out of my absolute mind.'
Call Her Daddy · The Unknowns of Motherhood
Recommendation · MoneyWill Brown recommends collecting traces from using agents; the traces contain your prompts and your feedback to things models did.
AI Engineer · Your Company Should Train Its Own AI Models - Will Brown, Prime Intellect
Recommendation · MoneyWill Brown recommends that a really useful starting point for an organization going from the abstract idea of training models on their data to actually getting there is evals as essentially test cases.
hedged
AI Engineer · Your Company Should Train Its Own AI Models - Will Brown, Prime Intellect
Recommendation · MoneyMolly Graham advises never shipping AI output straight to your boss without iterating or editing it — the way you would never forward an intern's presentation untouched. Treating AI output like it came from the best employee ever leads to copy-paste-send, which produces AI slop.
Lenny's Podcast · Molly Graham: The grief, burnout, and opportunity hiding inside the AI transition
Recommendation · MoneyMolly Graham recommends acknowledging that AI is not a super-intelligent being — it is much more like an intern. Literally think of it as an intern: ask what you would give to an intern and what you would never give them. She frames it as a tool to get work done while recognizing the work is still done mostly by humans and human brains.
hedged
Lenny's Podcast · Molly Graham: The grief, burnout, and opportunity hiding inside the AI transition
Recommendation · MoneyMolly Graham says the biggest skill to build right now is asking yourself 'can AI help me with this?' before doing any task. She compares it to meditation's stimulus-response gap: it creates space between the task and your response, letting you leverage AI tools in new ways. She also says to always think about what's the best use of your time, and then what an agent or robot could do for you.
Lenny's Podcast · Molly Graham: The grief, burnout, and opportunity hiding inside the AI transition
Recommendation · MindGreene advises asking yourself what you hate, what repels you, what repulses you — this reveals your life's task as much as what you love — and asking what the things are that you can't do, the ones that when you do them just feel wrong. He illustrates with his own experience: he hated working for other people (the egos, political games, and focus on charming the boss rather than results), never lasted more than 11 months at any single job in his life, and knowing he could not work for other people, the only profession he could pursue was writing books — the one profession without a boss.
Andrew Huberman · Access the Sublime & Live Your Best Life | Robert Greene