Working with AI · A field guide

From prompt to output: what decides how good AI work gets

How AI systems like Claude actually work, and the seven things that set the quality of what you get back.

The labThe modelWorkspaceConversationPromptSkills & agentsYour review

Before we start

Four ways to work with AI, and what makes one an agent

Chat

Chat + tools

AI workflow

Agent

Who decides the next step?

You, every turn

You; it may grab one tool to answer

Whoever designed the workflow

The AI, working towards your goal

Steps per request

One reply

One reply, a quick lookup first

A fixed sequence, the same every time

As many as the goal needs

Checks its own work?

No

Rarely

Only where checks are built in

Yes: tests, re-reads, fixes, retries

Best for

Questions, quick drafts, thinking out loud

Answers that need current facts or one document

High-volume, repeatable tasks: tagging, routing, summaries

Open-ended, multi-step work with files, data and tools

Example

“Suggest 5 subject lines”

“What did competitor X announce this week?”

Every new support ticket gets classified and summarised

“Audit our 40 landing pages and fix the off-brand copy”

Where you'll meet it

Claude.ai, ChatGPT, Gemini

The same apps with search or uploads on

Zapier, Make, n8n, the API in internal tools

Claude Code, Cowork, cloud agents

Agent= model+ tools+ context+ a loop that runs until the goal is met

Agents · the nuances

“Agent” isn't on or off. It's a matter of how much you let go.

Ask every stepIt proposes each edit or action; you approve. Slow but safe.

Plan firstIt researches and writes a plan; nothing changes until you OK it.

Auto-accept editsIt works freely inside the folder and asks before anything risky.

Runs on its ownIn the cloud or on a schedule; you review the result afterwards.

← more controlmore autonomy →

Same model, different setup

An agent isn't a smarter AI. Its power comes from tools, context and the loop. Give it a vague goal and you get a busy agent, not a good result.

One agent or a team

A main agent can hand parts of the job to helper agents (subagents), each with its own fresh context, then combine their findings.

Watching vs. in the background

Interactive: you watch and steer as it works. Background: it runs in the cloud or on a schedule and reports back when done.

New ways to fail

Small mistakes can compound over many steps, it can drift off-task, or take an action you didn't intend. Checkpoints and permissions are the remedy.

Define “done”

The single best agent instruction: what finished looks like, where it goes, and what's off-limits. That's what it checks its work against.

Don't use an agent when…

…a quick question or rewrite is all you need (chat is faster), or the process never changes (a fixed workflow is more predictable).

Three nuances matter most. First, autonomy is a dial. In Claude Code you choose: approve every step, plan first, auto-accept edits, or let it run in the background. Start on the left for anything unfamiliar or risky, and move right as trust builds. Second, the agent is not a different, smarter brain. The same model becomes an agent when it gets tools, context and a loop, which is why everything later in this talk (workspace, prompt, skills) matters even more for agents. Third, agents fail differently: a chat gives you one bad answer, while an agent can take twenty steps in the wrong direction. Hence: define done, use checkpoints, and set permissions. Finally, agents aren't always the answer. For a one-line question, chat is faster; for a process that never changes, a fixed workflow is cheaper and more predictable.

Agents · Multi-agent setups

When one agent isn't enough: five ways to use several

Delegate

Subagents

The main agent hands a focused job to a helper with a fresh context. Only the conclusion comes back.

Cost

e.g. A researcher reads 40 review sites and returns one page of themes.

Parallelise

Fan-out

One job split into independent pieces that run at the same time, then merged.

Cost

e.g. Five competitors analysed at once, merged into one comparison table.

Review

Maker and checker

A second agent critiques the first one's work against the brief, the facts or the brand rules.

Cost

e.g. A writer drafts the launch email; a brand and fact checker flags issues before you see it.

Collaborate

Agent team

A lead plus teammates who share a task list and message each other directly, including challenging each other. Experimental in Claude Code.

Cost

e.g. Customer view, competitor view and devil's advocate debate the Q1 positioning.

Orchestrate

You run several

You start separate sessions, locally or in the cloud, one per piece of work, and check in on each.

Cost

e.g. One session per market localising the same campaign.

What you gain

Clean context in each agent, speed from parallel work, and independent viewpoints that catch blind spots.

What it costs

Each agent has its own context, so usage multiplies. Plus coordination overhead, detail lost in summaries, and clashes if two edit the same file.

Rule of thumb

Start with one agent. Add helpers when the work splits cleanly or needs a fresh pair of eyes. For teams, 3–5 members is plenty.

Five patterns, roughly from cheapest to most expensive. Subagents are the everyday one: Claude Code spawns them on its own to keep the main conversation clean. Fan-out is subagents in parallel, great when the pieces don't depend on each other. Maker and checker is underrated: a model reviewing its own draft in the same conversation tends to approve it, while a separate agent with a fresh context and a critic's brief finds real problems. Agent teams go further: teammates are full sessions that message each other and share a task list, which suits debate and research, but they're experimental in Claude Code, off by default, and use significantly more tokens. Finally, the simplest multi-agent setup of all is you running several sessions side by side. The trade-off is always the same: more agents means more speed and perspectives, but more cost and coordination. Start with one.

The whole journey

Where you run it decides what the AI can use

① Workspace

Chat history

Memory

Standing instructions

Your files

Skills

Connectors

Plugins

Subagents

Acts on files

Runs without you

Where its setup lives

ChatClaude app on web, desktop, phone

Your account settings

Claude Projectchats sharing files & instructions

Account + project instructions & files

Coworkdesktop agent on a folder

Account + installed plugins + the folder

Claude Code · localon your computer

Global ~/.claude + the folder’s .claude

Claude Code · cloudon Anthropic's servers

The repo’s .claude, not your laptop

Inside other appsChrome, Excel, Slack…

That app’s integration

built inlimited or opt-innot availableColumns 1–4: what it can see · columns 5–10 (③ tools): what it can do

2

Your promptsame in every row (next slide)

→

4

ModelFast · Workhorse · Frontier

→

5

The AI worksreads, plans, acts, checks

→

6

Outputtext, file, action

→

7

You review↺ back to ②

Read this top to bottom: each row is a place you can run a prompt, and each one switches on more of the AI's abilities. A plain chat sees the thread plus whatever you've opted into in your account settings (memory, skills, connectors), but it can't touch your files. A Project adds shared instructions and reference files. Cowork, in the Claude desktop app, works on a folder on your computer and adds plugins and helper agents. Claude Code is the fullest toolkit: it reads your global setup in ~/.claude on that computer plus the folder's own .claude settings. In the cloud it runs on Anthropic's servers from a GitHub repository, so it uses the repo's settings, not your laptop's, and it keeps working after you close the lid. The right-hand column answers "where do my skills and plugins come from?". Gotcha: skills added in the Claude app don't automatically appear in Claude Code, and vice versa. The prompt and the model are the same choices in every row, shown in the strip below.

② The prompt

What goes into a prompt depends on the job you're asking for

Every prompt mixes: Context+ Task+ Goal+ Constraints+ Output format … but each job leans on different ingredients.

Job

Example

Must include

Add this line

Ask

“What do the new EU pay-transparency rules require of us?”

Your situation, and what you'll use the answer for

Cite sources. Say if you're unsure.

Create

“A launch post for CFOs about automated approvals”

Audience, goal, tone (ideally an example), length

Ask me questions before you start.

Transform

“Turn this 20-page report into a 1-page exec summary”

The source material, the new reader, what must survive

Keep every number unchanged.

Analyse

“What's driving churn in this survey export?”

The data, the question, the decision it informs

Show the evidence for each finding.

Think

“Pressure-test our Q1 positioning”

Constraints, options already considered, how to judge

Give 3 options with trade-offs.

Do

“Research 5 competitors and save a comparison file”

Definition of done, where to save, what's off-limits

Plan first; wait for my OK.

The big picture

Output quality is a stack, and you control most of it

01

The lab

Which company built the AI. It sets the ceiling on capability, safety and data handling.

IT / procurement

02

The model

Which model in the family, and how much “thinking” it's allowed.

You

03

Workspace context

Standing instructions, reference files, memory: what the AI knows before you ask.

You / your team

04

Conversation context

The thread so far. It can build understanding, or pile up noise.

You

05

The prompt

What you ask, how precisely, and in what order you ask it.

You

06

Skills, tools & agents

Reusable playbooks, connections to company data, and helpers that can take actions.

You / your team

07

Your review

Checking, critiquing and steering. The step that turns a draft into finished work.

You

Under the hood

A language model does one thing: predict the next piece of text

1 · Text is split into tokens

Roughly ¾ of a word each. Models read, think and bill in tokens.

Our Q3 campaign targets finance leaders at mid-size firms who are frustrated by

3 · Repeat, token by token

Every word of the answer is produced this way. The model has no memory between chats and no view of your company except what's included in the text it's given.

2 · It scores every possible next token

…frustrated by ___ (illustrative probabilities)

manual

34%

slow

21%

expense

15%

spread

10%

their

7%

…50k more

13%

Everything you put in front of the model shifts these odds. That's the whole reason context and prompts matter.

The context window

Each turn, the model reads one big bundle, and your prompt is a small part of it

System

Tools & skills list

Workspace

Conversation history

Files read & tool results

You

Free space

Illustrative split, mid-way through a typical Claude Code session. Modern windows hold a few hundred thousand tokens or more: several books' worth.

System prompt. The app's built-in rules and personality.

Tools & skills. Short descriptions of what the AI can do.

Workspace. Your standing instructions and saved memory.

History. Every message so far, yours and the AI's.

Material. Documents it opened, searches, data it pulled.

Your prompt. Often under 5% of what's read.

Top labs leapfrog each other every few months. Pick a lab for its platform, not just for this month's lead.

Data & compliance

Where data is processed and kept, and whether it's used for training. In a corporation, this usually decides the lab for you.

Closed vs. open weights

Closed: used through the lab's app or API. Open: you can download the model and run it in-house, but you also run and secure it yourselves.

Layer 1 · How they compare

For writing and professional work, the latest scores

Creative writing: which output people prefer

Arena (LMArena): blind side-by-side votes from real users. Best model per lab · Elo · axis starts at 1380.

Professional deliverables (incl. marketing)

GDPval-AA (Artificial Analysis): real tasks from 44 occupations (docs, decks, spreadsheets), judged head to head. Best model per lab · Elo · axis starts at 1450.

Close races are ties. On writing, Gemini 4 Argon and Claude Opus 5.5 are within the margin of error.

Snapshots, not verdicts. Rankings shift every few weeks. Use them as a shortlist.

Your own test beats any chart. Run 5 of your real tasks through 2–3 models and compare.

Data as of 2 Oct 2026: arena.ai/leaderboard/text/creative-writing · artificialanalysis.ai/evaluations/gdpval-aa

Layer 2 · The model

Three model tiers: pick one for the job, then set how hard it thinks

Fast

Haiku 4.5

Capability

Speed

Cost

Quick, cheap, good enough for simple, high-volume jobs.

Tagging feedback, short rewrites, summarising emails, quick lookups

Other labs: “Flash”, “mini”, “Lite” models

Workhorse

Sonnet 5.5

Capability

Speed

Cost

Near-frontier quality for most tasks at a fraction of the price. The right default.

Drafting copy, documents, slides, spreadsheets, research summaries

Other labs: their standard model

Frontier

Opus 5.5 · Fable 5.1

Capability

Speed

Cost

Deepest judgement and nuance. Opus tops the writing and pro-work rankings; Fable is a premium tier for very long autonomous work.

Strategy, positioning, long-form content, ambiguous problems

Other labs: “Pro”, “Max”, flagship models

The second dial: effort / thinking

The same model can answer straight away or reason first. More effort means better answers to hard problems, but it's slower and costs more. Turn it up for analysis and strategy, down for quick edits.

Rule of thumb

Start on the workhorse. Move up when the output lacks judgement or nuance, and down when you're doing the same simple thing hundreds of times.

Claude model line-up as of October 2026. Dots show relative position within the family.

Layer 5 · The prompt

Anatomy of a strong prompt

You're a senior B2B content strategist.¹ We're launching automated expense approvals for finance leaders at companies with 500–5,000 staff. Their pain: month-end close drags on because approvals sit in inboxes.² Write a LinkedIn post announcing it.³ The goal is demo bookings, not likes.⁴ Keep it under 150 words, use one concrete number, avoid words like “revolutionary”, and end with a question.⁵ Here's a past post that performed well, for tone: <example>…</example>⁶ Give me 3 versions with different hooks (data-led, story-led, contrarian) in a table, with a one-line rationale each.⁷ Before writing, ask me up to 3 questions if anything important is missing.⁸

1

RoleSets the expertise and point of view.

2

Context & audienceThe part people skip most, and the one that matters most.

3

TaskOne clear verb and deliverable.

4

Goal / the “why”Lets it make smart trade-offs you didn't spell out.

5

ConstraintsLength, must-haves, words to avoid.

6

ExampleShows the tone faster than any adjective.

7

Output formatNumber of options, structure, layout.

8

ProcessHow to work: ask first, plan first, or check its own draft.

② The prompt · Write it once

Write the context once, and every prompt gets shorter

Without a workspace

Monday · LinkedIn post~130 words

Tuesday · trial email~130 words

Wednesday · video script~130 words

Role, audience, voice, rules, examples and format retyped every time, or forgotten and left to guess.

Consistent: every task, and every teammate using the folder, starts from the same brief.

Rule of thumb: said it twice? Move it to the file, or just ask “add this to CLAUDE.md”.

In the Claude app: a Project does the same with instructions and knowledge files.

→

With CLAUDE.md and reference files

CLAUDE.md read every session

## Audience
Finance leaders, 500–5,000 staff. Pain: slow month-end close.

## Voice
Senior B2B strategist. Direct, no hype. Never “revolutionary”.

## Defaults
Social posts <150 words, one concrete number.

## Delivery
3 variants in a table, save to drafts/.

## How to work
Ask up to 3 questions if the brief has gaps.

brand/voice-guide.mdtone, with good & bad examples

examples/top-posts.mdpast winners to imitate

research/personas.pdfwho the buyers are

product/facts.mdfeatures, pricing, claims we can make

Read when relevant; CLAUDE.md points to them.

Your prompts now carry only what's new · ~15 words

MonLinkedIn post announcing automated approvals. Goal: demo bookings.

TueEmail to trial users who haven't set up approvals yet.

WedTurn Monday's post into a 30-second video script.

Layer 5 · Word choice

Specific words change the output; vague words leave it to chance

Swap vague words for checkable ones

make it engaging→open with a surprising number about month-end close

keep it short→under 120 words, 3 short paragraphs

professional tone→like a CFO writing to a peer: direct, no hype

for our audience→for finance directors who already use an ERP

don't be wordy→one idea per sentence, plain verbs

make it better→make the benefit clear in the first line

Say what to do, not just what not to do. Naming a banned phrase can plant it.

Phrases that change how it works

Ask me questions before you startSurfaces the context you forgot to give

Think this through before answeringMore reasoning on hard or multi-step problems

Give me 3 options that differ in…Breadth instead of one safe answer

Be critical. What's weak here?Counters its natural agreeableness

Only use the attached documentsKeeps it grounded in your sources

If you're unsure, say soFewer confident-sounding guesses

Cite where each claim came fromMakes fact-checking fast

Here's why this matters: …Better judgement on things you didn't specify

Layer 5 · Best practices

Prompting habits that consistently pay off

01

Brief it like a smart new hire

Brilliant, but on day one. What would they need to know to do this well?

02

Explain the why

“Readers skim on mobile” beats “use short paragraphs”, and it carries over to choices you didn't spell out.

03

Show, don't describe

One or two real examples of “good” say more than a paragraph of adjectives.

04

Material first, question last

With long documents, paste the material at the top and put your ask at the end. Label each part clearly.

05

Define the output

Length, structure, number of options, table or prose. Don't make it guess the shape.

06

One job per message

Big tasks go better as a sequence of steps than one giant request.

07

Let it interview you

“Ask me questions one at a time until you have enough to write this.” Great for briefs and strategy.

08

Separate draft from critic

After a draft: “Now review it as a sceptical CFO.” Then: “Fix the top 3 issues.”

09

Ask it to improve your prompt

“Here's my prompt. What's ambiguous or missing?” Prompt-writing is a task it's good at.

10

Edit, don't argue

If a reply goes wrong, edit your earlier message and re-run instead of piling corrections on top.

11

Save what works

A prompt that worked becomes a template, then a project instruction, then a skill.

12

Spot-check against reality

Numbers, names, quotes and links are where errors hide. Verify those first.

Layer 5 · Sequencing

Good work comes from a sequence of prompts, not a single one

1

Brief

Give the situation, audience and goal. Hold back the writing.

Here's the launch, the audience and our goal. Don't write yet. What do you need to know?

2

Explore

Get options before committing. It's cheap to discard ideas here.

Suggest 4 different angles, with the trade-offs of each.

3

Draft

Commit to one direction with a clear choice.

Go with angle B. Draft it, keeping the constraints from earlier.

4

Critique

Switch the AI from writer to tough reviewer.

Review this as a sceptical CFO. List the 5 weakest points.

5

Revise

Make targeted fixes, not a full rewrite.

Fix points 1, 2 and 4. Leave the rest unchanged.

6

Package

Turn the result into the formats you'll actually use.

Final version, plus 3 subject lines and a 2-line summary for my manager.

Start a fresh chat when…

you switch topics, it keeps repeating the same mistake, or the thread has got very long.

…and carry the essentials over

“Summarise what we've decided and the constraints, as a brief I can paste into a new chat.”

Not every task needs all six

A quick rewrite is just step 3. A strategy doc deserves all six, possibly several times over.

Layer 4 · Conversation context

Everything in the thread stays in play, including the wrong turns

Draft a post for our new expense feature.

“Revolutionise your finance team!” …

No, way too hypey. Also it's for CFOs.

“CFOs: revolutionise your month-end!” …

Still hypey. And drop the exclamation marks.

“Month-end close shouldn't wait on an inbox.” …

The crossed-out turns are still read on every later turn. Their hype keeps pulling the output back.

When context helps

Decisions, preferences and corrections build up. By turn 10 it knows your audience, tone and constraints without being told again.

When context hurts

Rejected drafts, abandoned ideas and unrelated topics compete for attention. Very long threads get summarised automatically, and detail is lost.

How to keep it clean

Edit & re-run the message that went wrong, instead of correcting after it.

One topic per chat. A new task gets a new thread.

Summarise & restart once a thread has done its job.

Layer 3 · Workspace context

The workspace briefs the AI before you've typed a word

📁 q4-launch/

├─ CLAUDE.mdbrand voice, audience, rules · read every session

├─ brand/

│ └─ voice-guide.mdreference · read when relevant

├─ research/

│ ├─ personas.pdf

│ └─ q3-results.csvdata it can analyse

├─ drafts/where it writes output

└─ .claude/

├─ skills/linkedin-post/a reusable playbook

└─ agents/fact-checker.mda helper agent

In Claude Code, the folder you open is the workspace. In the Claude app, Projects work the same way: instructions plus knowledge files.

Loaded every session

CLAUDE.md files: yours, the team's and the folder's standing instructions
Memory: facts it saved from earlier sessions
Names & one-line descriptions of available skills, agents and connectors

Loaded only when needed

Files it opens or searches
The full instructions of a skill that matches the task
Data fetched through connectors

Write the brief once, in CLAUDE.md, and every prompt in this folder starts from it.

Layer 6 · Capabilities

Tools, skills, connectors, plugins and hooks: what each one adds

Tools

Its hands

Built-in actions: read and write files, search, browse the web, run commands.

read · edit · search · web fetch

Skills

Playbooks

A folder of instructions, templates and examples for one kind of task. Picked up automatically when the task matches.

brand-voice · board-deck · pdf

Connectors

Plugs into company systems

Via MCP, a standard connection protocol. Lets the AI read from and act in other apps.

Drive · Slack · Jira · CRM

Plugins

A role-in-a-box

One install that bundles skills, agents, connectors and commands for a role or team.

marketing · sales · design

Hooks

House rules that always run

Automatic checks at fixed moments. They run every time, so the AI can't forget them.

spell-check after each edit · block deletes

Why many skills don't clutter the context: they load in stages

Always: name + one-line description (a few dozen tokens)

→

When the task matches: full instructions

→

If needed: templates, examples, scripts

③ Tools · Beyond text

Claude Code can drive other tools, and chain them into workflows

Three ways in: connectors (MCP), APIs called from a short script, and command-line tools. Each one needs your company's approval and its own account or key.

Campaign asset factory

From one brief: headline and copy variants, on-brand images in 4 ad sizes, all saved and named in a folder and sent to Slack for review.

Brief.md→Claude→Image API→Slack

Monday performance digest

Pulls last week's traffic and ad spend, spots what moved and why, makes charts and posts a one-page summary. Runs on a schedule.

GA4+Ads→Claude→Charts→Notion

Localise a campaign

Adapts the master campaign for 6 markets (not just translation), then generates a native-sounding voiceover for each video cut.

Master copy→Claude→Voice API→Drive

Competitor watch

Visits competitor pricing and feature pages weekly, compares them with last week's snapshot and flags any change worth reacting to.

Web→Claude→Diff file→Email

Personalised outreach

Takes a CRM segment, writes tailored email variants per persona and loads them back as drafts. A human approves before sending.

HubSpot→Claude→HubSpot drafts

Design hand-off

Reads a Figma frame, writes copy that fits the actual space, checks it against the brand guide and fills it back into the layout.

Figma→Claude+Brand guide→Figma

Layer 6 · Agents

An agent is a model working in a loop, with tools and a goal

Chat

You ask, it answers once. You do the legwork between turns.

Agent

You set a goal. It does the legwork, step by step, then reports back.

A real run in Claude Code

Goal: “Compare our pricing page with our 3 main competitors and draft positioning ideas.”

GatherReads CLAUDE.md and brand/positioning.md

ActFetches 3 competitor pricing pages from the web

PlanOne page failed to load, so it tries the cached version

ActWrites drafts/pricing-comparison.md with a table

CheckRe-reads the table and fixes a mismatched price tier

DoneSummarises 3 positioning angles and flags 2 assumptions to verify

Layer 6 · Agents, continued

Subagents: helpers with their own fresh context

Clean context

The heavy reading happens in the helper's window. Only the conclusion comes back, so the main thread stays sharp.

Parallel work

Several helpers can run at once: research, analysis and checking happen side by side.

Specialists

Each subagent can have its own instructions, tools and even a different model, like a cheap fast one for bulk reading.

Layer 7 · Your review

You stay accountable: verify, steer and protect

Verify

Check where errors actually happen

Numbers, dates, names, quotes
Links and sources: open them
Claims about your own company or product
Anything it couldn't actually have seen

AI can be fluent and wrong at the same time. Polished text isn't proof.

Steer

Control what agents may do

Ask first: approve each edit or action
Plan mode: it researches and proposes, you approve before it acts
Auto-accept: only for low-risk, reversible work
Interrupt and redirect at any time

Protect

Handle company data properly

Use the company-approved tool and plan (enterprise plans don't train on your data)
No customer personal data or secrets in personal accounts
Connectors inherit your access, so be deliberate
Label AI-assisted work where policy requires it

The AI drafts; you decide. Your name goes on the output, so your judgement has to go into it.

Troubleshooting

When output disappoints, find the layer that failed

Symptom

Usual cause

Fix

Generic, “AI-sounding” text

Prompt No audience, goal or example

Add the context and the why; paste one example of good work

Ignores earlier instructions

Conversation Thread too long or cluttered

Summarise and start fresh; put standing rules in project instructions

Wrong about our company

Workspace It was never told

Add reference files and a CLAUDE.md or project instructions

Confident but invented facts

Review Not grounded or checked

“Use only the attached sources and cite them”; verify the specifics

Shallow analysis

Model Too small or too little effort

Switch to a bigger model or turn up thinking

Same fix needed every time

Skills Know-how not captured

Turn the working prompt into a skill or template

Agent went off-track

Agent Vague goal, no checkpoints

Use plan mode, define “done”, and ask it to check its own work

The compounding loop

Don't just fix the output. Fix whatever produced it.

If the output……fix it here

Sounds off-brand→Add a brand-voice reference file with good and bad examples

Gets facts about us wrong→Add a reference file: products, pricing, positioning

Forgets a house rule→Write the rule into CLAUDE.md or the project instructions

Has the wrong structure→Put a template inside the skill

Needs the same 5 corrections→Turn the corrections into a skill

Lacks live data→Add a connector to the source system

Breaks a must-never rule→Add a hook: an automatic check that always runs

Fix the draft by hand

Helps once. Same problem next time.

Fix the system

Helps every time after, and for your whole team.

Take this with you

Eight habits for better AI output, starting Monday

Model

Match the model to the stakes

Workhorse by default; flagship for judgement calls.

Workspace

Write the brief once

Standing instructions in a project or CLAUDE.md.

Conversation

One topic, one thread

Edit instead of arguing; restart with a summary.

Prompt

Context and the why

Audience, goal, example and format beat clever wording.

Sequence

Brief, explore, draft, critique

Several focused steps beat one mega-prompt.

Skills

Capture what works

Good prompts become templates, then skills.

Agents

Delegate goals, not keystrokes

Give an agent a clear “done” and checkpoints.

Review

Trust, then verify

Check the specifics. You own the output.

The model is the engine. The context you give it is the steering.

References

Sources & further reading

Benchmarks (data as of 2 Oct 2026)

Arena: Creative Writing leaderboard · arena.ai/leaderboard/text/creative-writing
Artificial Analysis: GDPval-AA · artificialanalysis.ai/evaluations/gdpval-aa

Prompting

Anthropic: Prompt engineering overview · docs.claude.com

Claude Code

Overview & getting started · docs.claude.com/…/claude-code/overview
Memory & CLAUDE.md · …/claude-code/memory
Skills · …/claude-code/skills
Subagents · …/claude-code/sub-agents
Plugins · …/claude-code/plugins
MCP connectors · …/claude-code/mcp
Hooks · …/claude-code/hooks

Questions?

From prompt to output: what decides how good AI work gets

Four ways to work with AI, and what makes one an agent

Chat

Chat + tools

AI workflow

Agent

“Agent” isn't on or off. It's a matter of how much you let go.

Same model, different setup

One agent or a team

Watching vs. in the background

New ways to fail

Define “done”

Don't use an agent when…

When one agent isn't enough: five ways to use several

Subagents

Fan-out

Maker and checker

Agent team

You run several

What you gain

What it costs

Rule of thumb

Where you run it decides what the AI can use

What goes into a prompt depends on the job you're asking for

Output quality is a stack, and you control most of it

A language model does one thing: predict the next piece of text

1 · Text is split into tokens

3 · Repeat, token by token

2 · It scores every possible next token

Each turn, the model reads one big bundle, and your prompt is a small part of it

A handful of labs build the frontier models

Anthropic

OpenAI

Google DeepMind

xAI

Meta

DeepSeek

Alibaba

Moonshot AI

Z.ai

Mistral

Capability

Data & compliance

Closed vs. open weights

For writing and professional work, the latest scores

Creative writing: which output people prefer

Professional deliverables (incl. marketing)

Three model tiers: pick one for the job, then set how hard it thinks

The second dial: effort / thinking

Rule of thumb

Anatomy of a strong prompt

Write the context once, and every prompt gets shorter

Without a workspace

With CLAUDE.md and reference files

Your prompts now carry only what's new · ~15 words

Specific words change the output; vague words leave it to chance

Swap vague words for checkable ones

Phrases that change how it works

Prompting habits that consistently pay off

Brief it like a smart new hire

Explain the why

Show, don't describe

Material first, question last

Define the output

One job per message

Let it interview you

Separate draft from critic

Ask it to improve your prompt

Edit, don't argue

Save what works

Spot-check against reality

Good work comes from a sequence of prompts, not a single one

Brief

Explore

Draft

Critique

Revise

Package

Start a fresh chat when…

…and carry the essentials over