the new jev model is insane.. it can check your AI's work and make decisions inside software for $0.042 per million input tokens, with free output here are the first things you should use jev for: 1. second brain 2. content workflow 3. post analysis 4. SEO article review typesafe reports up to 193.6× faster results and 444.6× lower costs than the LLMs in its workflow tests you give jev information and specific questions. it returns choices, scores or probabilities that your software uses to decide what happens next 1. second brain when a document, message or meeting note enters your second brain, give jev the content and your categories identify what it is, classify the topic and check for duplicates against existing notes your software checks required fields, saves the content with its source and verifies that it was saved correctly works well with the karpathy LLM wiki framework 2. content workflow give jev your draft, and then reference your anti slop rules and voice DNA file (which defines how you write) it will check for generic phrasing, repeated points and differences from your voice. use previous content and its results to estimate performance potential then send results to the writing agent for revision, compare the performance estimates with results after publishing 3. post analysis jev can compare a draft with previous posts and their results to estimate how well it could perform with your audience evaluate the hook, topic and format against defined criterias. and compare those scores with the performance after it goes live 4. SEO article review give jev your article / page draft, a target keyword and the articles currently ranking for that keyword (in the top 10 SERP) compare how well they answer the query, cover the topic and provide useful information. use that comparison to estimate your article's ranking potential then your LLMs can revise weak sections before publishing, and compare the estimate with actual rankings so many more usecases, we will see many new upgrades to previous concepts and workflows we´ve read about now

Every way to earn on Creator Wire just got a home. Introducing Opportunities: one page showing every active deal and way to make money on the platform > Invites for paid deals > Collabs for brand badges > Boosts for reward pools > Referrals for your own link > Challenges for judged contests. > Viral launch campaigns Every reward pool sits in escrow before it opens, and every payout lands as real money. The page shows your own numbers, matched to what you can earn right now, so you know exactly where you stand before you post a single thing. Explore it ↓

how to use Jev to check your ai marketing team's work before you review it jev is a new model built to evaluate information and make structured decisions inside software you give it context and specific questions, and it returns choices, scores and probabilities. that makes it useful for checking your marketing agents' work before it reaches you for review when your agents create a landing page, social posts or an email sequence, you can have the work checked against the campaign brief and your requirements an eval is a check against a requirement you've defined. for marketing, that could mean checking for ai tells, unsupported claims or confusing copy here's how to set that up: 1. agree on the requirements with your lead agent start with your campaign goal, brief and the assets you're creating have the agent ask for your experience, previous feedback and anything you already know needs attention. include the mistakes you keep correcting when reviewing work for a landing page, define checks for: > ai tells in the copy, like forced contrasts, repetitive phrasing and generic filler > invented facts, quotes, customer results or unsupported claims > whether product details and other information match your approved sources > confusing headings, unclear buttons or missing information someone needs to take the next step > visual issues like cramped spacing, weak contrast, inconsistent typography and broken mobile layouts give the evaluator examples of work you've approved and rejected, with your reasons. turn "this looks generic" into specific things it can check save those requirements in the campaign brain before the creating agent starts work jev can assess the copy and supplied evidence. pair it with browser tests and visual inspection for the ux/ui checks, then send the combined findings back to the creating agent before your review 2. give the evaluation step the right context your company brain holds reusable knowledge about the business, including positioning, voice, supported claims and exclusions the campaign brain holds the brief, research, decisions and working files for the campaign have your workflow retrieve the relevant files and send them to jev alongside the asset and evaluation questions if you're checking a customer result, include the case study that supports it. for writing style, include your banned patterns and examples with your edit notes have the workflow request missing material before evaluating the draft 3. ask specific questions jev can evaluate jev takes context and questions with defined answer formats it can choose from options you provide, score something against criteria or estimate the probability that a statement is true. independent questions can run in parallel against the same context for your landing page, that could look like: > does this paragraph use any of the banned writing patterns? > does the case study support the result quoted in this paragraph? > do the product details match the supplied documentation? > does the button label explain what happens when someone clicks it? your workflow uses those answers to decide what happens next, and the creating agent handles any revisions 4. send failed checks back with enough context to fix them your orchestrator coordinates the work and assigns tasks to the agents connect the evaluation results to its revision process: > failed checks return to the agent responsible for the asset > uncertain results go to your review queue with the issue flagged > assets that pass reach your sign-off queue with their evaluation attached if a customer claim fails, send the copy agent the specific claim, the failed requirement and the supporting material it was evaluated against after the revision, run the checks on the updated asset again set a limit on revision attempts. anything that keeps failing should reach you with the unresolved issue and what the agent has already tried 5. include technical and visual checks you can use this process across your marketing work, with different tools handling different checks code can check character limits, required fields and tracking parameters. browser tests can check forms, links and how a landing page renders designs and videos need tools that can inspect their visual content, along with your review. jev evaluates the information you supply, so confirm which formats it supports before assigning it checks on those assets keep the results attached to the asset, including anything that wasn't checked also run checks as work progresses. research should be evaluated before it informs the brief, and claims should be checked before they get reused across the campaign 6. test the checks against work you've already reviewed take previous assets you've approved and rejected, including your edit notes run the evaluation against them and compare its decisions with yours look closely at drafts that pass even though you would reject them. you may need to make a requirement more specific, provide missing context or change how the question is asked jev's confidence estimates also need testing against your examples before they control whether work continues automatically keep the results advisory during testing, with a person reviewing the decisions 7. save what you change during review when an asset reaches you, review the work alongside its failed checks, revisions and anything still unresolved you still decide whether the angle makes sense, whether the creative feels right and whether it's ready to publish save the reasons for your edits in the campaign brain if you keep correcting the same issue, turn it into a proposed evaluation requirement. test it against previous assets before adding it to the workflow keep campaign-specific feedback with that campaign

how to set up marketing agents that go out and get the context your campaigns need when you're planning a campaign, your agents can research competitors, explore marketing campaigns and signup funnels, read customer conversations and follow what's happening on the channels you're using give them the tools to access those sources and a process for turning what they find into a campaign knowledge you can use for planning, execution and experiments 1. start with what you need to understand give the agent your campaign idea and goal / objective, then have it read the relevant company knowledge and previous campaigns your company brain holds reusable knowledge about the business. the campaign brain holds the context, decisions and working files for this campaign have the agent ask about your experience before planning the research. you may have already tested an approach or learned something that never made it into a file work together to define the questions the research should answer: > which alternatives are customers comparing us with? > what objections should we address? > how are similar offers being presented? > which formats could suit this audience and channel? > what do we need to understand before choosing an angle? include the audience, market, language and how recent the information needs to be 2. connect the tools and write the research skill your agent needs tools for finding sources, reading pages, inspecting visual experiences and retrieving data connect web search, a page reader, browser or computer access, and relevant data apis through your agent environment test each connection with a small task and check that the agent can retrieve the content and keep its source then write a research skill explaining: > how to turn the brief into research questions > which sources and tools to use > what evidence to collect > where to save the material > when to ask you for direction > when it has enough evidence to return set request and spending limits, with approval required for purchases, account changes or form submissions 3. research marketing campaigns and signup funnels give your codex or grokbot setup browser access to explore how other companies attract users and guide them toward signing up have it follow the public journey from an ad, social post or search result through the landing page and signup flow ask it to collect: > the campaign hook, creative and offer > how the landing page follows up on the message that brought someone there > how pricing, proof and customer objections are presented > the calls to action and steps leading to signup > what information someone is asked to provide > links and visual references for ideas you could test review the examples together and connect useful findings to the relevant pages in your campaign brain 4. research customer language and objections send the agent to relevant trustpilot reviews, reddit communities, category-specific review sites, app stores and comment sections give it questions to investigate, like why customers switch products, what disappoints them and what they wish they had understood before buying keep exact phrases with their sources and dates. have the agent group recurring themes while preserving disagreements and differences between customer types review those findings against your experience, customer calls and support tickets before using them to shape the campaign 5. connect search data and investigate customer queries use an seo data api to retrieve keyword data and search results for the market you're targeting research agents can investigate different groups of queries, such as product comparisons, customer problems and buying questions record the query, location, language and collection date alongside the findings you can also have different search-enabled models investigate those questions and collect the sources they find. when testing how ai products answer customer questions, save the answers and cited sources separately from keyword demand data the lead agent reviews the findings across these tasks and identifies what needs further investigation 6. check trends daily and weekly give the research agent a watchlist based on your audience, category, competitors and active campaigns connect the campaign to a grok bot or grok cli with x search enabled to investigate relevant keywords, accounts and conversations. add browser access where you want it to inspect timelines and visual posts for daily checks, have it look for developments that could affect work already underway: > conversations involving your audience > competitor launches or offer changes > emerging hooks and creative formats > questions your campaign could help answer > relevant events and creator activity use the weekly review to compare findings over time and decide which deserve an experiment each proposed opportunity should include the sources, why it fits your audience and what you could test 7. ingest the research into your campaign brain keep original pages, screenshots, transcripts and data in research/raw/, with their sources and collection dates have the agent review that material and ingest useful findings into wiki/, the campaign brain this follows the karpathy method, his llm wiki approach, applied to a campaign: the agent reads raw sources and builds a wiki of linked markdown pages that it maintains as new information comes in as it learns, it creates or updates linked markdown pages about customers, competitors, offers, objections, campaign examples and experiments those connections form a knowledge graph the agents can navigate a customer objection might link to a competitor's offer, a landing-page decision and the experiment you're running to test it. each page also links back to the evidence behind it have the agent check existing pages before creating new ones, update findings when evidence changes and record conflicting evidence or unanswered questions alongside the relevant topic the campaign brain also holds the brief, decisions, ideas and working documents, with links to designs and performance data kept in other tools 8. review the findings and use them in experiments build evals into the research skill, meaning checks against the requirements you agreed check whether sources support the claims, quotes are exact, dates fit the brief and findings apply to your audience then work through the recommendations together. your experience may change how the research should be used, and that reasoning belongs in the campaign brain for each experiment, record: > the finding that prompted it > what you expect to happen and why > the asset or campaign element you will change > how you will evaluate the result > what needs your sign-off before execution agents creating copy, designs and other assets can follow the wiki's links to the relevant findings, evidence and agreed decisions

Introducing Higgsfield API. 50+ frontier models in one API, at lower prices than a subscription. > Get up to 50% OFF discount on your 3 favorite models > Lock in your max-discount within 7 days > Pay per use with no commitment Build your own Higgsfield with the best prices in GenAI industry. Available at

how small businesses can use ai to handle incoming leads over the phone when you're running ads, someone can find your business and call while the team is busy with customers or has already gone home if phone calls are part of your funnel, it's worth checking what happens to those inquiries a voice agent can answer questions about your services, check availability, book an appointment and send a confirmation text here's how to set it up around your marketing 1. give it the context behind the call start with your website, then add the details someone responding to your campaign is likely to ask about > the service and offer you're advertising > pricing, eligibility and the areas you serve > common questions and objections > what information you need before booking if your ad promotes a specific offer, make sure the agent can explain the conditions 2. decide when it should answer you might want your team to take calls first during office hours, with the agent handling calls when nobody picks up or after you've closed define what it can book and which requests need someone from your team 3. test it like part of the campaign call with the questions you expect from a lead, including vague requests, questions about the offer and requests for a person check whether the answers are accurate and the booking details are correct 4. review what callers are asking recurring questions can show you what to explain on the landing page or in the ad. use those conversations to update the agent's information and improve the campaign ElevenLabs just launched Reception for this, an ai receptionist platform for small businesses built on ElevenAgents it handles incoming calls, answers questions, books appointments and sends confirmation texts, with setup starting from your website

how to set up marketing agents that go out and get the context your campaigns need when you're planning a campaign, your agents can research competitors, explore marketing campaigns and signup funnels, read customer conversations and follow what's happening on the channels you're using give them the tools to access those sources and a process for turning what they find into a campaign knowledge you can use for planning, execution and experiments 1. start with what you need to understand give the agent your campaign idea and goal / objective, then have it read the relevant company knowledge and previous campaigns your company brain holds reusable knowledge about the business. the campaign brain holds the context, decisions and working files for this campaign have the agent ask about your experience before planning the research. you may have already tested an approach or learned something that never made it into a file work together to define the questions the research should answer: > which alternatives are customers comparing us with? > what objections should we address? > how are similar offers being presented? > which formats could suit this audience and channel? > what do we need to understand before choosing an angle? include the audience, market, language and how recent the information needs to be 2. connect the tools and write the research skill your agent needs tools for finding sources, reading pages, inspecting visual experiences and retrieving data connect web search, a page reader, browser or computer access, and relevant data apis through your agent environment test each connection with a small task and check that the agent can retrieve the content and keep its source then write a research skill explaining: > how to turn the brief into research questions > which sources and tools to use > what evidence to collect > where to save the material > when to ask you for direction > when it has enough evidence to return set request and spending limits, with approval required for purchases, account changes or form submissions 3. research marketing campaigns and signup funnels give your codex or grokbot setup browser access to explore how other companies attract users and guide them toward signing up have it follow the public journey from an ad, social post or search result through the landing page and signup flow ask it to collect: > the campaign hook, creative and offer > how the landing page follows up on the message that brought someone there > how pricing, proof and customer objections are presented > the calls to action and steps leading to signup > what information someone is asked to provide > links and visual references for ideas you could test review the examples together and connect useful findings to the relevant pages in your campaign brain 4. research customer language and objections send the agent to relevant trustpilot reviews, reddit communities, category-specific review sites, app stores and comment sections give it questions to investigate, like why customers switch products, what disappoints them and what they wish they had understood before buying keep exact phrases with their sources and dates. have the agent group recurring themes while preserving disagreements and differences between customer types review those findings against your experience, customer calls and support tickets before using them to shape the campaign 5. connect search data and investigate customer queries use an seo data api to retrieve keyword data and search results for the market you're targeting research agents can investigate different groups of queries, such as product comparisons, customer problems and buying questions record the query, location, language and collection date alongside the findings you can also have different search-enabled models investigate those questions and collect the sources they find. when testing how ai products answer customer questions, save the answers and cited sources separately from keyword demand data the lead agent reviews the findings across these tasks and identifies what needs further investigation 6. check trends daily and weekly give the research agent a watchlist based on your audience, category, competitors and active campaigns connect the campaign to a grok bot or grok cli with x search enabled to investigate relevant keywords, accounts and conversations. add browser access where you want it to inspect timelines and visual posts for daily checks, have it look for developments that could affect work already underway: > conversations involving your audience > competitor launches or offer changes > emerging hooks and creative formats > questions your campaign could help answer > relevant events and creator activity use the weekly review to compare findings over time and decide which deserve an experiment each proposed opportunity should include the sources, why it fits your audience and what you could test 7. ingest the research into your campaign brain keep original pages, screenshots, transcripts and data in research/raw/, with their sources and collection dates have the agent review that material and ingest useful findings into wiki/, the campaign brain this follows the karpathy method, his llm wiki approach, applied to a campaign: the agent reads raw sources and builds a wiki of linked markdown pages that it maintains as new information comes in as it learns, it creates or updates linked markdown pages about customers, competitors, offers, objections, campaign examples and experiments those connections form a knowledge graph the agents can navigate a customer objection might link to a competitor's offer, a landing-page decision and the experiment you're running to test it. each page also links back to the evidence behind it have the agent check existing pages before creating new ones, update findings when evidence changes and record conflicting evidence or unanswered questions alongside the relevant topic the campaign brain also holds the brief, decisions, ideas and working documents, with links to designs and performance data kept in other tools 8. review the findings and use them in experiments build evals into the research skill, meaning checks against the requirements you agreed check whether sources support the claims, quotes are exact, dates fit the brief and findings apply to your audience then work through the recommendations together. your experience may change how the research should be used, and that reasoning belongs in the campaign brain for each experiment, record: > the finding that prompted it > what you expect to happen and why > the asset or campaign element you will change > how you will evaluate the result > what needs your sign-off before execution agents creating copy, designs and other assets can follow the wiki's links to the relevant findings, evidence and agreed decisions

how external research becomes part of your marketing agents' knowledge when an agent researches competitors, customer conversations or formats for a campaign, some of what it finds will be useful again a recurring customer objection might change how you explain your offer and this week's trending format might only be relevant for the social posts you're creating now part of context management is deciding what to fetch for the current project and what your agents should keep using across future campaigns 1. separate company context from campaign work your company brain holds the context you reuse across campaigns, like what you sell, who your customers are, your voice and the decisions you've agreed to keep each campaign has its own campaign brain, a folder for its goal, brief, research, decisions and all the working files, including ideas, drafts, documents and results it also links to designs and other work kept in external tools, so the agent can find everything connected to that campaign agents use the campaign brain alongside the relevant company files while working external research starts in the campaign brain. when you and the agent identify something worth using across future campaigns, you add that finding to the company brain with its source and your reasoning 2. work out what you need to learn start with your campaign idea and goal / objective, then give the agent access to the context you have have it read relevant previous campaigns and ask about your experience before planning the research. you may have already tested an approach, ruled out an audience or learned something that never made it into a file 3. fetch research that helps with those decisions depending on the campaign, that could include: > competitors' current pricing, offers and positioning > how others approached similar campaigns and launches > trends and formats on the channel you're publishing to > comparable landing pages, ads, thumbnails and hooks > customer questions and objections in comments, reviews and relevant discussions > search results and ai answers for the questions your customers are asking > creators and potential partners for the campaign > seasonal moments and buying triggers, like hiring or funding keep the channel, audience and date attached to each finding. a format getting attention on x gives you something to investigate for x, with separate research needed for linkedin 4. save the research with the campaign use campaigns//research/ for the material collected for that project keep the original sources, then have the agent write the relevant findings into research/distilled. md with links back to the evidence include when the research was collected, why each finding matters and anything that still needs checking references. md links to this research alongside the company files, performance data and design references needed for the campaign large collections can stay in the research files or database. have the agent search and retrieve the relevant material as it works through each task 5. bring your experience into the decisions when research affects the campaign direction, have the agent explain its recommendation and ask for your context before proceeding you might recognize an objection from customers you don't serve, or know that a popular format previously brought in the wrong audience work through that together and save the agreed decision with its reasoning in the campaign brain. anything still waiting on your judgment goes into the sign-off queue this applies throughout the campaign, including changes made while creating and reviewing the assets 6. add lasting lessons to the company brain once a finding repeats, has support from your own results or becomes a decision you've signed off on, review whether future campaigns should use it > recurring customer phrases and objections go into customer. md > competitor findings that affect your positioning go into competitors. md > formats that performed well for you go into formats. md, with approved work in examples/ > decisions you want future campaigns to follow go into rulings. md, including why you made them keep the source, date and conditions with each addition. a format that worked for a particular offer and audience should carry that context into the next campaign

context management is one of the core skills for a marketing engineer when you give an agent a campaign to work on, it needs to understand the business, use what you've learned from previous campaigns and follow your taste in the work it creates your second brain gives it the company knowledge, but you also need to connect it to performance data and a brand book you can work in together tldr > company brain = helps the agent understand your business and the work you're doing > warehouse = lets the agent measure performance and learn from results > brand book = guides the agent's design choices using your style and feedback keep references. md in the project so whichever harness you're using can find the relevant files, query the data and open the designs with your feedback 1. connect your context sources with references. md your harness is the software that runs the agent, like cursor, grok bot, claude code or hermes references. md holds links and instructions for finding context. the company files stay in the brain, performance data in the warehouse and designs in the brand book before making a decision, the agent should ask for your relevant context and experience, including what you've already tried, what you learned and anything that hasn't been written down yet for a landing page + social posts, the agent needs to: > work with you to define the campaign brief, starting from your idea, goal and experience > read company knowledge, previous campaigns, decisions and lessons from the brain > query previous campaign performance in the warehouse > open the brand book and approved designs, including your feedback > bring the evidence back to you and work through decisions together save the agreed brief and the reasoning behind your decisions in the campaign brain as the work progresses 2. keep company knowledge and decisions in the brain use markdown in git, with github as the shared place to keep it. the agent reads the files relevant to the job your company brain holds context that should remain useful across campaigns > what you sell, who buys and why > positioning, supported claims and exclusions > voice-dna. md, hooks. md and approved work > dated decisions and permissions to edit, spend or publish your campaign brain holds the context for the work in progress > this campaign's goal, audience, offer and constraints > decisions you've already made and why > approved versions, rejected drafts and your edit notes keep performance data in the warehouse. the visual/ folder holds links to design. md and the live brand file 3. keep performance in the warehouse use postgres or another sql database. define what each row represents, such as one asset in one campaign on one date keep the facts you use to make marketing decisions > spend, impressions, clicks, leads and conversions > content results, like saves, replies or watch time > lead source and stage in your customer relationship management system > experiments, dates and what you stopped running keep definitions beside the numbers. define what counts as a lead or conversion, which reporting window to use and how much the data can lag use shared identifiers like campaign_id and asset_id to connect records across sources recurring jobs pull the data on a schedule. the warehouse connects the records and calculates totals with sql, the language used to query the database your agent queries the warehouse for the results it needs, instead of pulling bulk history from each platform into its context for the landing page job, it can query the last similar campaign by asset over the last 28 days, using the same conversion definition save your interpretation in the campaign brain with a link to the query. keep the performance history in the warehouse 4. keep your design rules and visual feedback in the brand book this needs both written rules and a live design file you can work in together design. md gives the agent instructions it can read before creating anything > typefaces, sizes and when to use them > colors, contrast and logo placement > layouts, safe zones and formats for each channel > do / don't examples with the reason beside them > what it can create and what it should only adapt > links to the paper or figma file and the pages to open write your brand guidelines in design. md and link each guideline to an example in paper or figma. include your feedback on why a design works or what needs to change paper and figma mcp give the agent access to the live design file. mcp is the connection that lets it use the design tool keep your brand book, moodboard and approved work there > add a note explaining what you like about each reference > keep approved ads, landing pages and social designs > keep rejected versions with your comments on the frames > leave feedback next to the work you're discussing design. md explains the requirements. the live file lets the agent inspect examples, create designs and work alongside you as you review them 5. use all three context sources, and save what decisions and changes for the landing page + social posts, the agent reads the relevant company files and campaign brief, queries previous performance, then reads design. md and opens the brand page with approved examples work through the direction together before drafting copy and frames in paper or figma. send the work to your sign-off queue for review after approval, save the copy and the reason for your edits in a campaign folder. keep the approved frames and design feedback in the brand file after launch, results go to data warehouse and your interpretation goes in the company / campaign brain the models are interchangeable, but your company knowledge, performance history and design judgment carry into the next campaign you´ll be working on

how to route your codex / GPT models for marketing use GPT 6 Astra for thinking, judgment and planning, then let Sol coordinate the work and route tasks downstream give Astra your campaign idea and goal / objective, then point it to the context you have → Astra · high reasoning works with you to think through how the campaign should be done, weigh different approaches and shape the direction. use it with Matt Pocock's /wayfinder skill to map the path to completion, with tickets, evals to assess the work and decisions you need to be involved in → Sol · medium reasoning turns the agreed plan into execution tasks, coordinates their order and gives each model the context it needs. if campaign copy depends on an approved offer, Sol waits for that decision before assigning the draft → Luna · high reasoning handles execution that needs extensive context and precision, like writing a landing page from customer interviews, product documentation and your brand guidelines → Terra · medium reasoning handles routine execution, like formatting approved copy for different channels, updating campaign records and applying specific edits Sol evaluates the outputs against the requirements and requests revisions. decisions that need further judgment go back to Astra before your sign-off

every marketing engineer needs a knowledge base, a second brain a central place for files that contain all the information, data, examples and instructions your agents need to produce quality work here is how to organize it for marketing: 1. start with the files every marketing team needs > company .md: what the business does and who it serves > customer .md: customer problems, buying triggers, objections and their own words > offer .md: what you sell, pricing, deliverables and promises you can support > positioning .md: why someone would choose you over the alternatives > voice .md: how you write, with examples and phrases to avoid > proof .md: case studies, approved claims and links to the evidence build these from material you already have: sales calls, proposals, customer interviews, website copy and previous campaigns for voice .md, include your edits. an approved post helps, but explaining why you changed the opening gives the agent something specific to apply 2. create a folder for each marketing function content, SEO, paid, creator marketing and email all need different instructions inside each folder, keep: > playbook .md: how the work gets done > examples/: approved work and rejected versions, with your feedback > tools .md: which tools and accounts agents can use > checks .md: what to check before returning the work > calendar .md: what's planned, in progress and ready for review reference the shared files when needed. if you change the offer, every team should be able to find the current version 3. organize content by platform an X post, a newsletter and a LinkedIn post need their own examples and writing guidance content/ ├── playbook .md ├── checks .md ├── tools .md ├── calendar .md ├── x/ ├── linkedin/ ├── newsletter/ └── visual/ inside each platform folder, save the formats you use, examples you've approved and drafts you've rejected 4. give agents visual references to work from keep a moodboard and brand book in Figma or Paper, with links in visual/ > moodboard: references for the look you want, with notes on what you like about them > brand book: colors, fonts, logos, layouts and rules for using them > approved examples: previous ads, landing pages and campaign assets give the agent access to those files when it creates visuals, and save your design feedback alongside the references 5. give each campaign its own folder campaigns/product-launch/ ├── brief .md ├── research/ ├── decisions .md ├── production/ └── results .md the brief holds the goal, audience, offer and constraints for that campaign decisions .md records what you've agreed on and why. if you reject an angle or change the offer, save the decision there agents working on the landing page, emails and social posts can then use the same approved direction 6. connect your performance data using a data warehouse a data warehouse stores performance data from your marketing tools in one place your agents can query connect your ad platforms, website analytics, sales tools and content accounts to collect spend, clicks, leads, conversions and campaign results the marketing folders can hold campaign reports and your interpretation of the results, with links to the data behind them for a new campaign, an agent can read the previous brief, inspect the assets and check how they performed include the reporting period and metric definitions so it can make a useful comparison 7. explain how agents should use the folders add an AGENTS .md file with instructions for working in the knowledge base > where to find the relevant files > which sources to trust > where to save drafts and research > how to flag conflicting information > which changes need your sign-off give the agent the files it needs for the task you're asking it to do 8. save what you learn while doing the work when you correct a draft, update the writing instructions or examples when a campaign reveals a new customer objection, add the evidence to customer .md when an offer changes, update offer .md and flag the active campaigns affected I'd keep the markdown files in GitHub so changes can be reviewed and traced back to whoever made them, with large creative files linked from shared storage start with one marketing function and a task you're already doing

1 domain expert + ai = 10x the output a big claim, but that's what I've seen when marketers become engineers when you already understand your market, why people buy and what makes a campaign work, learning to build opens up so much you can turn the work you've done for years into workflows that agents can run, then use your experience to review what comes back and push it further research that you'd struggle to fit into a campaign can happen alongside production. ideas that would've stayed in a document can become landing pages, lead magnets and campaign tests as you repeat the workflows, you improve the instructions, connect more tools and create evals around the mistakes you keep running into this is why I'm so excited about marketing engineering there's so much experience inside marketing teams that we can now apply across more work if you've spent years learning a marketing vertical, you already have something valuable to build with

we're becoming 10x the marketers we used to be the amount and the quality of the work we can produce as marketing engineers is honestly incredible building these systems lets agents handle more of the research and production, while we shape the campaigns and push the quality further agents can keep working through loops and scheduled workflows, with us reviewing what they bring back and steering the decisions we can take on more ambitious campaigns and explore ideas we previously wouldn't have had the capacity to pursue this is what gets me excited about marketing engineering we're learning engineering and applying the same principles software teams have used for years to marketing work we already understand it sounds a bit silly until you're doing it yourself and realize how much you can get done and honestly, we're in such a bubble when your feed is full of people building agent workflows, it's easy to feel like everyone is already doing this I think we're still EARLY, especially in marketing over the next few years, I expect marketing engineering to become a much bigger role inside companies as it gets easier to build products, knowing how to get them in front of the right people gives you something valuable to build a career or a company around

how to do marketing with GPT 6 Astra give it the context, data and tools to work on your marketing here is the setup: 1. knowledge layer (company / second brain) keep your company information, audience research, offer, brand voice and past campaigns in a markdown wiki the agent can read 2. data warehouse store your marketing data in a postgres database so agents can check campaign performance and compare results 3. data source connectors connect the tools you use for analytics, ads, sales and content so agents can pull fresh data 4. a good harness the app you use to work with the agent, manage its tools and review its work, @orca_build and @herdrdev working well for me, but native codex works as well 5. curated skills reusable instructions for workflows and specific jobs. a marketing copywriting framework for the writing, and skill bundles like @mattpocockuk to test for planning and building 6. design workflow keep a design database that agents can both read and write to, for me that has become @paper, I´m able to prompt agents to design based on design. md reference file and by visual input, astra 6 is pretty good at doing this 7. experience in a marketing vertical you need enough experience in something like SEO, paid ads or content to judge output and give useful feedback then you prompt it with a specific outcome, or dictate the prompt with something like Wispr Flow I find it faster to explain what I want, add context and give feedback as the work develops that could be creating a campaign with a landing page and lead magnet, using your company information and past campaign results run /goal loops, where the agent keeps working toward the outcome, and review the work as it develops change the angle, push back on the copy, ask it to check an assumption against the data your marketing experience gives you the judgment to steer those decisions taste matters

agent tools need an OpenRouter layer this could be a HUGE infrastructure opportunity in AI OpenRouter gave developers one place to discover models, call them through the same interface, compare prices and route requests by price, speed and availability one campaign agent might use Firecrawl for research, Stripe for payments, HubSpot for CRM and Slack for communication connecting those providers directly means managing separate accounts, credentials, billing, permissions and tool formats (mega headache) the opportunity is one agent-tool router that manages how agents discover and run actions across those providers: > describe the outcome > discover the right provider > connect the correct company account > see the price and permissions > run the work and verify the result > retry, send it for human review or fall back to another provider this goes beyond an MCP directory. MCPs gives agents a common way to discover and call tools, while the agent-tool router still needs provider success rates, prices, permissions and verified outcomes we are moving from APIs that expose individual product actions toward outcome endpoints an outcome endpoint lets an agent request a finished job while the provider handles the planning, execution, validation and delivery (basically: give the agent the job, get back the finished result + receipt) an agent needs a small set of jobs it can request: > research this market > reconcile this account > launch this campaign > resolve this support ticket > produce this report > update this forecast Firecrawl lets an agent request a research job, while Parallel takes a research task and returns a synthesis with sources Composio and Pipedream aggregate large tool catalogs and handle parts of discovery, account authentication and execution. the official MCP Registry provides shared metadata for public tool servers. x402 is a pay-per-call protocol that lets agents purchase access to compatible endpoints the complete agent-tool router is still missing (we have most of the pieces, nobody seems to own the full route yet) this is where it gets messy: tools are harder to route than models. a model request usually returns text, media or embeddings, while an agent tool can publish a post, refund a payment, contact a customer, delete data or change the state of a company account so the router has to know a lot more than "which endpoint is cheapest": > per-user identity and permissions > human sign-offs for spend and external actions > protection against duplicate writes, plus a way to undo partial changes > logs, receipts and checks that confirm the outcome > routing based on price, speed, reliability and quality > fallback only when providers can safely produce the same result one connection should handle discovery, credentials, approval and access rules, billing and execution, while every agent receives only the account access and actions required for its job for a marketing agent, that could mean selecting a research provider, connecting the client workspace, showing the expected spend, waiting for sign-off before publishing, then returning the live links and receipts to the campaign brain, the files holding the campaign plan and results people use human interfaces to configure the system, approve sensitive actions, monitor work and handle exceptions. agents call outcome endpoints to run approved jobs the open problem is proving the selected provider finished the job before the router accepts the result, retries or safely falls back

a company skill library needs 5 layers keep every approved skill in one shared library, agents check it at session start or on a schedule, then update their local copies automatically 1. source of truth > one shared GitHub repository > every approved change is committed to its history > every skill has an owner, version and last update 2. discovery > a searchable internal page shows what exists > each skill explains the task it runs, when to use it and the output it produces > people search by outcome, team or workflow 3. loading > the agent reads the skill when the task matches > it also reads the relevant company files, brand rules and examples > the skill defines the steps, checks and output format. the company files provide the facts 4. improvement > someone runs the skill on a campaign, client brief or report > they fix the missed step, weak output or instruction about which source to trust > the correction goes back into the shared version > permitted people and agents receive it when their files update 5. governance > changes are reviewed before they reach everyone > old versions stay recoverable > overlapping skills get merged, unused ones get retired > a permissions file defines which teams and agents can read restricted skills

every creator / influencer campaign I have seen starts with a spreadsheet after running 1000s of these at @LunarStrategy, the same problem started to show up when you start a new campaign, context from past ones slowly disappears. which creators converted, what briefs worked, what payment terms they agreed to, all of it lives in old sheets nobody opens we solved this with Creator Wire here is what it does: > aggregates all your campaigns in one place, past and current > tracks which creators you worked with, what they delivered, and what it cost > finds new creators in your niche based on who already performed > manages payments and briefs without jumping between tools > easily share campaign performance with stakeholders we built it with creators in mind > track every deal in one place (and find new ones) > get briefed properly instead of through scattered DMs > collect your stake instantly when the campaign is done