Tools
Agents and Skills
The agents and skills that ship free in Foothold, MilUX's open operating system for defence founders. What they are, how they help, where each came from, and the licences.
Agents and skills are how you get expert work out of AI without being the expert yourself. The ones on this page ship free inside Foothold, MilUX’s open operating system for defence founders, and they run on your own machine. This page explains what they are, how they help, and lists every one, with where it came from and its licence. We can work with you to tailor these, and many more that we use, to your specific use case.
What they are
A skill is a repeatable procedure, written as plain instructions that Claude follows the same way every time. You ask for the thing (size my market, draft a DPIA, check whether I am eligible to bid), the skill runs the steps in order, and you get a usable result instead of a blank page. Skills are Markdown files in your own vault. You can read them, change them, and they keep working with no server and no account in the loop.
An agent is a configured instance of Claude, set up to do a job on your behalf using those skills. MilUX runs its own business on a stack of agents, and Foothold is how we hand the same way of working to founders. The catalogue below is skills today; agents are coming to Foothold next.
How they help
They turn a procedure an expert would charge for into something you can run yourself in minutes: a tender response built around the real evaluation criteria, a market sized properly, an investor deck red-teamed before a real investor sees it, or your contacts captured so nothing leaks away between meetings. You get a usable result quickly, and it stays yours.
We check skills before we trust them, and we will check yours. Every skill is a set of instructions an agent runs with real access, so an untrusted one is a way to run hostile instructions against your data. We built the Skill Safety Audit to check any skill or agent before it is installed, and it ships free in Foothold too. If you find a skill or an agent you like the look of and want a second opinion, send it to us. We are happy to audit it, and to pass on the good ones so everyone benefits. Email matt@milux.co.uk.
The catalogue
Everything here ships free in Foothold, grouped by what it helps with. Open a category to see its skills, and open a skill for the problem it solves and an example of what it could do set up for you. Each links to its folder in the public Foothold repository, where you can read exactly what it does before you run it.
Business, bids and compliance7 skills
Bid Response
Problem
Defence bids are won or lost on the published evaluation criteria, yet most founders write what they want to say and get marked down on compliance and evidence.
What it does
Structure and draft a response to a defence tender, DASA/UKDI competition, or framework call-off, built around the opportunity's own published evaluation criteria. Runs after eligibility-check returns a GO.
Set up for you
Reads a live DASA or framework call, builds the compliance matrix, and drafts every answer against that call's exact scoring, with your evidence slotted in.
View in FootholdBuild a DPIA
Problem
If you handle personal or operational data you need a DPIA, but founders rarely know where to start and stall before they begin.
What it does
Build a Data Protection Impact Assessment (DPIA) for the founder's business from their Foothold vault, walking them through any unknowns. The skill reads what the vault already knows (business context, CRM, tools, lawful basis), asks about the tools they use and the types of personal data they hold,
Set up for you
Pulls the facts from your vault, walks you through the gaps, and produces a DPIA ready to show a buyer or the regulator.
View in FootholdCyber Essentials Ready
Problem
Buyers expect Cyber Essentials, yet the controls read as IT jargon to a non-technical founder setting up their own laptop.
What it does
Walk a non-IT person through configuring their personal Mac or Windows computer to meet the technical controls expected for UK Cyber Essentials (Montpellier question set). Scope is the user's primary work computer.
Set up for you
Checks your actual laptop against each control and walks you, setting by setting, to a pass.
View in FootholdEligibility Check
Problem
Founders burn weeks on bids they were never eligible to win, on cyber, clearance, export or commercial grounds.
What it does
Pre-bid go/no-go gate for a defence opportunity. Walks the user through the four walls that disqualify SME bids before quality is even scored: the cyber floor, the clearance floor, the export-control position, and commercial basics.
Set up for you
Weighs an opportunity against your cyber, clearance, export and commercial position and returns a clear go or no-go in minutes.
View in FootholdLate Payment Reminder
Problem
Chasing a late invoice is awkward and easy to put off, so cash that is owed sits unclaimed.
What it does
Draft a late-payment reminder email for an overdue invoice under UK statutory rules. Triggers when the user says things like 'invoice <number> is overdue, value £X', 'draft a late payment chase for <invoice>', 'chase <client> for invoice <number>', 'this invoice is late', or 'send a reminder for <in
Set up for you
Tracks which invoices are overdue and drafts the reminder, adding the statutory interest and escalating the tone as you chase.
View in FootholdRed-Team Investor
Problem
Founders fall for their own deck and only hear the hard questions when a real investor is already saying no.
What it does
Critique any investor-facing content from the point of view of a sceptical, experienced investor, so you see the weaknesses before a real investor does. Works on any form: a pitch deck, a one-pager, an executive summary, a full business plan, a financial model, a cold email or intro, an investment m
Set up for you
Tears into your deck as a sceptical investor, lists the questions you will be asked, and shows you where the story breaks.
View in FootholdWrite SOW
Problem
A vague scope invites scope creep and IR35 risk, but founders find a tight, outcome-based Statement of Work hard to write.
What it does
Turn a client conversation into an outcome-based, outside-IR35-shaped Statement of Work, with clearly defined deliverables, measurable acceptance criteria, a risk register, and a pricing recommendation. Use this skill whenever the user says any of: "write a SOW", "draft a statement of work", "turn t
Set up for you
Turns a client call into an outcome-based, outside-IR35 Statement of Work with deliverables, acceptance criteria, a risk register and a price.
View in Foothold
Product, market and strategy34 skills
Ansoff Matrix
Problem
Founders chase every growth idea at once, without weighing which are safe bets and which are risky leaps.
What it does
Map growth options on the Ansoff Matrix: market penetration, market development, product development and diversification, by risk and timeline, with prioritised strategies per quadrant. Use when weighing growth paths or planning expansion.
Set up for you
Maps your real product and market onto the grid and ranks the growth moves by risk and timeline.
View in FootholdBeachhead Segment
Problem
Trying to sell to everyone at once, a small firm wins no one; the hard part is choosing where to start.
What it does
Choose the first beachhead segment to win: the smallest winnable, referenceable purpose-segment that proves the offering and opens adjacent ones. Evaluates segments against acute need, ability to contract, winnability against incumbents and primes, and referral into adjacent programmes.
Set up for you
Weighs your candidate customers and names the smallest winnable segment to prove the product first.
View in FootholdBusiness Model Canvas
Problem
Founders hold the business model in their head in pieces, so the gaps and contradictions stay hidden.
What it does
Generate a Business Model Canvas (Strategyzer) with all nine blocks: partners, activities, resources, value propositions, customer relationships, channels, segments, cost structure, revenue streams. Use for an established business, corporate strategy, or investor materials where you need how all the
Set up for you
Builds your Business Model Canvas from what you tell it, and flags where the model does not yet hold together.
View in FootholdCompetitive Battlecard
Problem
In a competitive pitch, founders get caught flat-footed on why a buyer should choose them over an incumbent.
What it does
Create a sales-ready competitive battlecard against one named competitor or incumbent: overview, quick comparison, where we win and where they win, objection handling, and win/loss patterns. Use when preparing for a bid, a capture, or a 'why not competitor X' conversation.
Set up for you
Researches a named prime and hands you a battlecard: where you win, where they do, and how to answer the objection in the room.
View in FootholdCompetitor Analysis
Problem
Founders either dismiss the competition or fear it, without a clear read of who is really in the space and where the gap is.
What it does
Map the competitive landscape: the direct competitors, incumbents and primes, their strengths and weaknesses, and where you can differentiate. Use when doing competitive research, building a bid or capture picture, or finding differentiation against an incumbent.
Set up for you
Pulls the players in your space from public sources and maps where the real gap for you sits.
View in FootholdCustomer Interview Script
Problem
Founders ask leading questions and hear what they want, so their research confirms the idea rather than tests it.
What it does
Create a structured customer or user interview script on Mom Test principles: warm-up, jobs-to-be-done exploration, probing, and wrap-up, with a note-taking template. No leading questions, no pitching, focus on past behaviour.
Set up for you
Writes a Mom-Test interview script for the exact people you want to learn from, with questions that will not lead them.
View in FootholdCustomer Journey Map
Problem
Founders see their product, not the customer's whole experience, so they miss where it actually breaks down.
What it does
Build an end-to-end customer journey map: the stages a customer or user moves through, the touchpoints, what they do, think and feel, where the friction is, and what to improve. Use when mapping a customer or user experience, finding friction in a buying or in-service journey, or improving onboardin
Set up for you
Maps your buyer's journey end to end and marks the pain points and the moments that decide the sale.
View in FootholdDesign the cheapest test
Problem
Founders build the thing to find out whether anyone wants it, spending months on an assumption they could have tested in days.
What it does
Design the cheapest test that could falsify a riskiest assumption, with the change-my-mind result set before the test runs. Turns a load-bearing belief into a small, fast, falsifiable experiment, names the metric and the sample, and pre-registers the decision each result triggers (proceed, pivot, or
Set up for you
Takes your riskiest assumption and designs the cheapest experiment that could prove you wrong this week.
View in FootholdEstimate Market Size (TAM, SAM, SOM)
Problem
Founders quote a huge market number with no working, and investors stop listening; sizing it properly is fiddly.
What it does
Size a market opportunity using TAM, SAM and SOM with top-down and bottom-up estimates, framed for defence as programmes and budgets rather than consumer total-addressable-market. Use when sizing an opportunity, scoping addressable spend, preparing an investment or bid case, or evaluating market ent
Set up for you
Sizes TAM, SAM and SOM for your product, top-down and bottom-up, with every assumption on show for an investor.
View in FootholdGTM Strategy
Problem
A good product with no route to market goes nowhere, and founders default to hoping word spreads.
What it does
Build a go-to-market strategy: route to market, channels, messaging, success metrics and a phased launch plan. Use when planning a launch, entering a new market, or building a GTM plan.
Set up for you
Builds your route to market: the channels, the messaging and the first moves to make.
View in FootholdIdeal Customer Profile
Problem
Founders sell to whoever will talk to them, spreading thin instead of nailing the customer they fit best.
What it does
Define the Ideal Customer Profile: the purpose-segment whose job the offering best fits, the customers most likely to select, value and stay, and who is explicitly not a fit. Use when defining an ICP, sharpening a target market, or prioritising sales and capture effort.
Set up for you
Reads your early users and defines the customer whose problem you fit best, so you stop spreading thin.
View in FootholdLean Canvas
Problem
An idea feels coherent in your head until you lay out the whole model and see what does not hold.
What it does
Generate a Lean Canvas (Ash Maurya): problem, solution, unique value proposition, unfair advantage, segments, channels, revenue, cost structure, key metrics. A fast hypothesis-testing tool for a new venture.
Set up for you
Fills your Lean Canvas and pressure-tests problem, solution and model on a single page.
View in FootholdMarket Segments
Problem
Founders treat the market as one blob, missing that different buyers are hiring them for different jobs.
What it does
Identify 3 to 5 purpose-named market segments: the distinct jobs different customers hire the offering for, each with its selection criteria, pains, fit and competitive picture. Use when exploring a market, prioritising target audiences, or evaluating new segments.
Set up for you
Splits your market into the three to five jobs different buyers are really hiring you for.
View in FootholdMonetization Strategy
Problem
Founders default to a day rate or a single price, leaving better revenue models unconsidered.
What it does
Brainstorm three to five distinct revenue models for an offering, each with audience fit, unit economics, risks, competitive position and a cheap validation experiment. Use when exploring how to make money from a product or service.
Set up for you
Proposes several revenue models for your offer and shows how each would play out.
View in FootholdOpportunity Solution Tree
Problem
Founders jump to a solution before understanding the opportunity, and build the wrong thing well.
What it does
Build an Opportunity Solution Tree (Teresa Torres): map a single desired outcome to customer opportunities, candidate solutions, and the experiments that would validate them. Use when structuring discovery, organising opportunities, or deciding what to build or test next.
Set up for you
Maps the opportunities and solutions under the outcome you want, so you build the right thing next.
View in FootholdPESTLE Analysis
Problem
Founders miss the macro forces, regulation, politics, economics, that can make or break a defence venture.
What it does
Run a PESTLE analysis of the macro environment: political, economic, social, technological, legal and environmental factors, rated by impact and likelihood, with strategic responses. Use for a defence-landscape read, market-entry assessment, or strategic planning.
Set up for you
Scans the political, economic, social, technological, legal and environmental forces around your venture.
View in FootholdPitch Prep
Problem
Founders cram everything into a pitch and lose the room; the hard part is cutting to the sharp version.
What it does
Help a founder build a short, sharp pitch for a demo day, investor meeting, or customer conversation. Walks the three-act pitch structure, stress-tests the idea, and points to the tools for building the deck.
Set up for you
Shapes a short, sharp pitch for a named audience and cuts everything that loses the room.
View in FootholdPorter's Five Forces
Problem
Founders misjudge how attractive their market really is, ignoring suppliers, buyers, substitutes and new entrants.
What it does
Run a Porter's Five Forces analysis of industry structure: competitive rivalry, supplier power, buyer power, threat of substitutes, threat of new entrants, and the resulting attractiveness. Use when assessing a market's dynamics or a structural position.
Set up for you
Maps the five forces in your market and shows where the profit and the pressure really sit.
View in FootholdPositioning Ideas
Problem
Founders sound like everyone else in the category, so buyers cannot tell why they are different.
What it does
Brainstorm differentiated positioning ideas for an offering: read the top competitors and the gaps they leave, then generate positioning statements with rationale, supporting messages and the advantage behind each. Use when developing or sharpening positioning.
Set up for you
Generates sharper positioning angles your competitors have not claimed.
View in FootholdPre-Mortem
Problem
Optimism hides the ways a bid or launch will fail, until it does; naming the risks up front feels unnatural.
What it does
Run a pre-mortem on a plan, bid, launch or mobilisation: imagine it has failed and work back to the real risks. Sorts risks into Tigers (real), Paper Tigers (overblown) and Elephants (unspoken), classifies Tigers as blocking, fast-follow or track, and writes action plans for the blockers.
Set up for you
Imagines your bid or launch has failed and works back to the risks worth heading off now.
View in FootholdPricing Strategy
Problem
Founders price on cost or gut, leaving money on the table or undercutting their own value.
What it does
Design a pricing strategy grounded in value, competitive position and willingness to pay: pricing models, tiers, value metric, price sensitivity and experiments. Use when setting or changing prices.
Set up for you
Designs pricing grounded in the value you deliver, not just a day rate.
View in FootholdPrioritisation Frameworks Reference
Problem
Everything feels urgent, so founders work on what shouts loudest rather than what matters most.
What it does
Reference guide for selecting and applying a prioritisation method. Leads with the economic methods, Cost of Delay and CD3 (Cost of Delay Divided by Duration), then covers Opportunity Score, ICE, RICE, Kano, MoSCoW and the rest with formulas and when-to-use guidance.
Set up for you
Picks the right prioritisation method for your situation and applies it to your backlog.
View in FootholdProduct Strategy Canvas
Problem
Founders have a product but no strategy, so the roadmap is a wish list with no logic underneath.
What it does
Build a product or service strategy on the 9-section Product Strategy Canvas: vision, segments, relative costs, value propositions, trade-offs, metrics, growth, capabilities, defensibility. Use when setting strategy for an offering or service line.
Set up for you
Builds a product strategy on the nine-section canvas, from vision through to the bets you are making.
View in FootholdProduct Vision
Problem
Without a vision worth following, a founder struggles to rally a team or convince an investor.
What it does
Brainstorm an inspiring, achievable and emotional product or service vision that motivates a team and aligns stakeholders, with three to five options and a recommendation. Use when defining or refining a vision, or aligning a team around a shared direction.
Set up for you
Drafts a product vision worth rallying a team and an investor around.
View in FootholdSentiment Analysis
Problem
Feedback piles up faster than a founder can read it, and the signal in it goes unused.
What it does
Analyse a body of user feedback at scale to surface segments, sentiment scores, jobs to be done and satisfaction patterns, with prioritised recommendations. Use when you have reviews, surveys, support tickets or social listening to synthesise.
Set up for you
Reads a pile of user feedback at scale and surfaces the segments, the sentiment and what to act on.
View in FootholdStakeholder Mapping and Engagement Plan
Problem
In a defence programme the players and their agendas are opaque, and founders engage the wrong ones.
What it does
Map stakeholders on a power/interest grid and build a tailored engagement plan per quadrant. Use when managing stakeholders on an engagement, planning a capture or bid, aligning a cross-functional effort, or planning campaign audiences.
Set up for you
Maps a programme's players by power and interest and builds an engagement plan for each.
View in FootholdStartup Canvas
Problem
Founders lack one view that ties strategy and model together, so decisions get made piecemeal.
What it does
Generate a Startup Canvas (Pawel Huryn): the 9-section Product Strategy Canvas plus a business model (cost structure and revenue streams), separating strategy from model. The recommended single artefact for a new product or venture that needs both strategic clarity and a business model.
Set up for you
Fills your Startup Canvas so strategy and model sit in one view.
View in FootholdStrategy Red-Team
Problem
A plan looks solid to the person who wrote it, and the assumptions it rests on go unchallenged until reality does.
What it does
Red-team a bid, capture plan, business case, roadmap or strategy by attacking its load-bearing assumptions before reality does. Steelmans each claim, then attacks it, ranks failure modes by impact, likelihood and cheapness-to-test, and returns the cheapest test and kill criteria for each.
Set up for you
Attacks the assumptions under your capture plan before the competition or the customer does.
View in FootholdSummarise Customer Interview
Problem
Research interviews pile up as raw transcripts, and the findings never get pulled out and used.
What it does
Summarise a research interview transcript into a structured synthesis: jobs to be done, current solution, what works and what does not, satisfaction signals, key insights and dated action items. Use when processing discovery interview recordings or transcripts.
Set up for you
Turns a raw research transcript into a structured synthesis: findings, quotes and what they imply.
View in FootholdSWOT Analysis
Problem
Founders assess their position loosely, without tying internal strengths and weaknesses to the outside forces.
What it does
Run a SWOT analysis: strengths and weaknesses (internal) against opportunities and threats (external), cross-referenced into actionable recommendations. Use for a fast strategic read on a position.
Set up for you
Runs a SWOT that ties your internal strengths and weaknesses to the outside forces.
View in FootholdUser Personas
Problem
Founders design for an imagined average user, not the real people who choose and use the product.
What it does
Build research-backed personas: the customer who selects and the user who operates, each with their purpose (job to be done), pains, gains and what they judge on. Use when building personas from research, interviews or survey data, or profiling the people inside a segment.
Set up for you
Builds research-backed personas: the person who chooses your product and the one who uses it.
View in FootholdUser Segmentation
Problem
A user base looks uniform until it is segmented by what people are actually doing with the product.
What it does
Segment an existing user base from feedback and usage data into 3 or more distinct behaviour- and needs-based segments, grounded in the job each group is doing rather than demographics. Use when analysing diverse user feedback, building a segmentation model, or finding hidden user groups.
Set up for you
Segments your users from usage and feedback by the jobs they are really doing.
View in FootholdValue Proposition
Problem
Founders describe features, not the jobs, pains and gains a customer actually cares about.
What it does
Design a value proposition with a customer using a 6-part jobs-to-be-done template (Who, Why, What before, How, What after, Alternatives), expressed as how well the offering fits what the customer is selecting on. Use when articulating why a customer should choose the offering, sharpening positionin
Set up for you
Designs your value proposition against a real customer's jobs, pains and gains.
View in FootholdValue Proposition Statements
Problem
Founders struggle to turn what they do into words a buyer immediately gets, across site, deck and outreach.
What it does
Turn a value proposition into ready-to-use statements for marketing, sales and onboarding, each addressing a segment, leading with the benefit, and naming the capability behind it. Use when writing campaign copy, pitch and bid messaging, or onboarding messages.
Set up for you
Turns your value proposition into ready-to-use lines for your site, your deck and your outreach.
View in Foothold
Relationships and CRM6 skills
Add contact
Problem
Contacts met at events and in meetings scatter across memory, email and LinkedIn, and get lost.
What it does
Add a new contact to the CRM. Pulls public information about the person, drafts a contact page, asks the user to confirm before writing.
Set up for you
Pulls a new contact together from their LinkedIn, your past emails and the web, and files a clean, linked page in your vault.
View in FootholdAdd event
Problem
The context of who was where and said what fades fast once an event is over.
What it does
Capture a new event into the vault. Logs attendance, attendees, organisations represented, and key threads.
Set up for you
Captures an event and everyone at it, linked to the people and organisations already in your vault.
View in FootholdAdd organisation
Problem
Founders track companies loosely and lose the thread on primes, buyers and partners over a slow sales cycle.
What it does
Add a new organisation to the CRM. Pulls public information about the company, drafts an organisation page, asks the user to confirm before writing.
Set up for you
Creates an organisation page pre-filled from public sources, ready for you to build the relationship on.
View in FootholdImport Relationships
Problem
A founder already knows useful people, but that network lives in their head and their inbox, not a usable CRM.
What it does
Bootstrap the vault CRM from relationships the user already has, whether they live in a CRM (HubSpot, Salesforce, Pipedrive, or any other), a CSV export, or only in the user's head. Connector-first: if the user's CRM has an MCP connector available, suggest connecting it and read directly.
Set up for you
Reads your LinkedIn connections and your inbox and bootstraps a working CRM from the network you already have.
View in FootholdMeeting Prep
Problem
Founders walk into meetings underprepared on who they are seeing and what matters to them.
What it does
Prepare a briefing for an upcoming meeting. Pulls relevant context from the vault and from any connected tools (calendar, CRM, email, chat).
Set up for you
Pulls a briefing for tomorrow's meeting: who you are seeing, their history with you, and what matters to them.
View in FootholdMobilise Engagement
Problem
A new engagement starts in a scramble, with scope, contacts and context half-captured.
What it does
Bootstrap a Customer Engagement during the scoping phase. Use when the user says 'mobilise <customer>', 'start scoping <customer>', 'set up the engagement with <customer>', 'onboard the <customer> engagement', 'begin mobilisation for <customer>', or asks to kick off a new piece of client work.
Set up for you
Stands up a new engagement, scope, contacts and context, so you start organised rather than scrambling.
View in Foothold
Operations4 skills
AI Readiness
Problem
Founders feel they should be using AI but have no honest read on whether their data and processes are ready.
What it does
A light, honest self-assessment of how ready the organisation is to put AI to work, across four lenses: data, processes, people, and governance. Produces a one-page snapshot with a suggested first move per lens, biased towards things this pack can immediately do.
Set up for you
Assesses your data and processes and tells you honestly where you are ready for AI and where to fix first.
View in FootholdProcess Map
Problem
How things get done lives in the founder's head, so nothing can be delegated or improved.
What it does
Map one business process through a structured interview and write it up as a process page: the steps, who does each, inputs and outputs, pain points, and a candidates section flagging what could be automated and what must stay human. The foundation skill of the ops wing; sop-capture and ai-readiness
Set up for you
Interviews you about how a job gets done and writes the process up clearly enough to hand over.
View in FootholdService Blueprint
Problem
Founders see the customer-facing part of a service, not the backstage work that makes it happen or fail.
What it does
Map a customer-facing service as a service blueprint: the customer's journey stage by stage against the front-stage actions they see, the back-stage work they don't, and the systems underneath, with moments of truth and failure points marked. A user-centred design method most small businesses have n
Set up for you
Blueprints a service you offer: the customer's steps and all the backstage work behind them.
View in FootholdSOP Capture
Problem
Without written procedures, quality depends on the founder being in the room every time.
What it does
Turn a mapped or described process into a runnable standard operating procedure: numbered steps a competent person could follow cold, explicit decision points, and a what-good-looks-like check at the end. Reads the process page from process-map where one exists.
Set up for you
Turns a process into a runnable SOP someone else could follow without you in the room.
View in Foothold
Content2 skills
Humaniser
Problem
AI drafts read as AI, and publishing that under your name costs credibility in a sceptical sector.
What it does
Remove signs of AI-generated writing from text so it sounds more human. Mandatory pass for any public-facing content before handover (LinkedIn posts, newsletter copy, web copy, press releases, external briefs, client proposals, case studies).
Set up for you
Strips the AI tells from any draft so it reads like you wrote it, before it goes out under your name.
View in FootholdNewsletter Writer
Problem
Founders know they should keep in touch with their audience, but writing each edition from scratch stalls it.
What it does
Draft a newsletter edition following the operator's brand voice and structure. Use when the user says 'write the next newsletter', 'draft an edition on [topic]', 'newsletter on [topic]', 'help me write a newsletter', or runs /newsletter-writer.
Set up for you
Drafts each newsletter edition in your voice from your week's updates, ready to edit and send.
View in Foothold
Brand1 skill
Design System Setup
Problem
Founders end up with inconsistent colours, type and logos across everything, and the brand looks amateur.
What it does
Build a complete, token-based design system for a founder's brand from a short interview: colour, type, spacing, a generated showcase page, and starter templates (PDF, slide, letterhead). Use when the user says 'set up a design system', 'create our brand system', 'build our design tokens', 'I need a
Set up for you
Becomes the engine to rapidly create branded artefacts and design your first website, all from one consistent set of colours, type and rules.
View in Foothold
Meta and maintenance7 skills
Curator
Problem
A growing vault quietly rots, with broken links and missing fields, and tidying it by hand never happens.
What it does
Budgeted vault hygiene sweep. Fixes mechanical defects silently within hard per-run caps (broken wikilinks with an unambiguous target, missing or malformed frontmatter, unescaped pipes in table wikilinks) and reports everything needing judgement to a dated Curation Report.
Set up for you
Sweeps your vault each month, quietly fixes broken links and missing fields, and reports anything that needs your judgement.
View in FootholdDaily Brief
Problem
Founders start the day unsure what matters, rebuilding context from scattered notes and calendars.
What it does
Write today's daily note as a brief FOR the user, generated from everything they are working on: engagements, initiatives, outstanding items rolled from previous notes, upcoming events and deadlines. The user reads the brief; they do not fill in a form.
Set up for you
Writes you a short brief each morning from your engagements, your tasks and today's meetings.
View in FootholdFoothold
Problem
Improvements to the pack are useless if updating risks your own work, so founders leave it stale.
What it does
Pull the latest content from the public Foothold GitHub repo into the user's installed pack and reconcile it with anything they've edited locally. Self-upgrades first if a newer version of the skill itself is on GitHub, so future skill changes propagate automatically.
Set up for you
Pulls the latest Foothold improvements into your vault safely, without touching your own work.
View in FootholdProcess Interviewer
Problem
Founders start building before the plan is clear, and the gaps only surface once they are expensive to fix.
What it does
Relentless one-question-at-a-time interviewer that extracts a complete, unambiguous plan before any work starts. Use when scoping an initiative, building a new skill, designing a workflow, or stress-testing your own thinking.
Set up for you
Interviews you one question at a time until a fuzzy idea becomes a complete, buildable plan.
View in FootholdPrompt Master
Problem
A vague prompt gives a vague result, and founders blame the tool rather than the ask.
What it does
Generate a production-ready prompt for a specific AI tool. Use whenever the user needs to write, fix, improve, or adapt a prompt for any LLM, coding agent, browser agent, or research agent — including Claude, Claude Code, Cursor, Cline, Copilot, GPT models, reasoning models (o3/o4-mini/Qwen3-thinkin
Set up for you
Turns a rough ask into a production-ready prompt for whichever AI tool you are using.
View in FootholdSkill Safety Audit
Problem
A skill you found online runs with real access to your data, and there is no obvious way to know whether it is safe.
What it does
Audit an agent skill or a pack of skills for safety before you install, run, or share it. Reads every file in the skill (SKILL.md, bundled references, assets, and scripts, not just the main instructions), looks for hidden or malicious instructions, data exfiltration, calls to unknown domains, attemp
Set up for you
Audits any skill you find before you install it and gives a clear safe-or-not verdict.
View in FootholdSkill Value Review
Problem
Faced with a pile of third-party skills, founders cannot tell which are worth using, forking, or ignoring.
What it does
Review a third-party agent skill, or a GitHub repo of skills, for value and fit, then recommend whether to use it as is, fork and adapt it, find an alternative, or build your own. Runs skill-safety-audit first and gates on the verdict, then scores the skill on a fixed value rubric and writes up stre
Set up for you
Reviews a skill, or a whole repo of them, and tells you whether to use, fork, replace or build your own.
View in Foothold
Something missing?
If you see an area that we’ve missed or want us to help you build an agent or skill, drop us a line.
Licences
Everything here is MIT licensed and free to use, change and pass on. Foothold itself is MIT.
The skills marked MilUX original are ours, released MIT under the Foothold pack. The forked skills keep their upstream copyright and licence, retained in full in each skill’s folder:
- The product, market and strategy skills are forked from Pawel Huryn’s pm-skills (MIT, © 2026 Pawel Huryn), some lightly adapted for defence founders.
- Humaniser is forked from @blader/humanizer (MIT), based on Wikipedia’s “Signs of AI writing”.
- Process Interviewer and Prompt Master are forked from Ben AI.
Provenance is shown against each skill above. If any attribution here is wrong, tell us and we will fix it.