General Business Skills

Posts 1–4 built the mechanics of skills and how they compose. This post is the first of two on the kinds of skills (general business here, domain-specific in post 6). General business skills matter because the bulk of most organizations’ repeatable work is not specialized research but ordinary, process-shaped tasks — and those are exactly what skills can make reliable at scale.

What general business skills capture

General business skills encode the common, recurring, process-shaped work that spans roles and organizations — the everyday tasks that are valuable, frequent, and done a particular way:

• Document production — drafting a standard kind of document to your conventions: a report, a proposal, a summary, a memo, meeting notes, a status update. The agent produces it in your format, with your structure and tone, every time.
• Process execution — following a defined business process: a review procedure, an approval workflow, an onboarding checklist, a standard request-handling flow. The skill encodes the steps so they’re followed consistently.
• Analysis and synthesis — performing a standard analysis: summarizing a set of documents, extracting key points, comparing options against your criteria, preparing a recurring analysis the way you always do it.
• Communication — drafting standard communications in your voice: responses, updates, explanations — consistent in tone and content with how your organization communicates.

What these share is that they’re broadly applicable, recurring, and convention-bound — not deep domain expertise, but the way your organization does ordinary work. They’re high-value precisely because they’re frequent: a skill that makes a daily or weekly task reliable and consistent pays off constantly, and the cumulative time saved across an organization’s ordinary work is enormous. General business skills are the “quiet majority” of a useful skill library.

The design challenge: too generic is useless

General business skills have a specific failure mode worth naming, because it’s the main way they go wrong: being too generic. The whole point of a skill (post 1) is to capture your specific way of doing the task — and a “write a report” skill that just says “write a good report” adds nothing the general model didn’t already do. The value is entirely in the specifics:

• Your conventions, not generic best practice. The agent already knows how to write a report; the skill must capture how you write your reports — your structure, your sections, your tone, your required elements, what you include and omit. The specifics are the skill; the generalities are noise (and context-wasting, post 2).
• Your process, not a generic process. “Review this” is generic; “check these specific things, in this order, against these criteria, escalating X to Y” is a skill. The particular steps and criteria your organization uses are what make it reliable and worth capturing.
• Your voice and constraints. Communication skills live or die on capturing your actual tone, terminology, and constraints — a generic “professional tone” skill is worthless; your real style and rules are valuable.

So the design discipline for general business skills is capture the specific, cut the generic (post 3’s craft, sharpened here): include exactly what’s particular to how you do the task, and nothing the model already knows. A good general business skill reads like your organization’s actual playbook for that task — concrete conventions, specific steps, real constraints — not a generic how-to. This is counterintuitive because the tasks feel generic (“just writing a report”), but the value is always in the organization-specific specifics, and skills that miss this are the most common disappointment.

Why these are high-leverage — and how to find them

General business skills are often the highest-leverage place to start a skill library, for reasons worth seeing:

• They’re frequent, so the payoff compounds. A skill for a task done many times a day across many people saves time and enforces consistency constantly — the leverage is in frequency × people, which for ordinary business work is large. The unglamorous daily task, made reliable, often beats a sophisticated rare one in total value.
• They enforce consistency across people. When everyone’s agent uses the same skill for “our standard report,” the output is consistent regardless of who’s driving — a real organizational benefit (consistency, quality floor, brand/voice coherence) that’s hard to achieve through training humans alone. Skills are a consistency mechanism, not just a speed one.
• They capture tacit organizational knowledge. Much of “how we do things” lives tacitly in experienced people’s heads (the institutional-memory problem). General business skills make that tacit know-how explicit and reusable — which both preserves it and spreads it, turning “ask Sarah how we format the quarterly deck” into a shared skill.

To find the general business skills worth building, look for the frequent, convention-bound, currently-inconsistent tasks: what do people here do repeatedly, a particular way, that varies in quality depending on who does it? Those are prime skill candidates — the recurring report, the standard review, the routine analysis, the common communication. Building skills for them captures the organization’s ordinary expertise, enforces its standards, and compounds value through frequency. General business skills aren’t glamorous, but they’re frequently where a skill library delivers the most total value — precisely because they make the vast ordinary work of an organization reliable, consistent, and fast.

The throughline: general business skills capture the common, recurring, convention-bound everyday work — document production (reports/proposals/memos to your format), process execution (reviews/approvals/onboarding your way), analysis/synthesis (your standard analyses), and communication (your voice/rules) — the “quiet majority” of a useful library, high-value because frequent. Their defining failure mode is being too generic: the model already knows how to write a report, so the value is entirely in capturing your specific conventions, process, voice, and constraints — capture the specific, cut the generic (a good one reads like your actual playbook, not a generic how-to). They’re high-leverage because frequency × people compounds the payoff, they enforce consistency across people (same skill → same output regardless of who drives), and they make tacit organizational knowledge explicit and reusable (institutional memory). Find them by looking for the frequent, convention-bound, currently-inconsistent tasks — “what do people do repeatedly, a particular way, that varies by who does it?”

Key takeaways

• General business skills capture common, recurring, convention-bound everyday work: document production (reports/proposals/memos/notes to your conventions), process execution (reviews/approvals/onboarding/request-handling your defined way), analysis/synthesis (your standard analyses), and communication (drafts in your voice) — the broadly-applicable “quiet majority” of a skill library, high-value because frequent.
• Their defining failure mode is being too generic: the general model already does a generic version, so a “write a good report” skill adds nothing — the value is entirely in your specifics (your structure/sections/tone, your exact process steps and criteria, your actual voice and constraints).
• Design discipline: capture the specific, cut the generic — include exactly what’s particular to how you do the task and nothing the model already knows (generalities are noise and waste context, post 2); a good one reads like your organization’s actual playbook, not a generic how-to (counterintuitive because the tasks feel generic).
• They’re high-leverage: frequency × people compounds the payoff (an everyday task made reliable saves time constantly), they enforce consistency across people (same skill → same output regardless of who drives — a quality floor and voice coherence training alone can’t easily achieve), and they make tacit organizational knowledge explicit/reusable (preserving and spreading “how we do things” — institutional memory).
• Find them by looking for the frequent, convention-bound, currently-inconsistent tasks: “what do people here do repeatedly, a particular way, that varies in quality by who does it?” — often the highest-total-value place to start a library, precisely because they make the vast ordinary work reliable, consistent, and fast.

Further reading

• Business process automation — automating recurring business work
• Standard operating procedure — the convention-bound processes skills capture