AICRO

Skill Library

Every AICRO-authored skill, from the repo tier (.claude/skills/) and the machine-wide global tier (~/.claude/skills/), grouped by execution tier and ordered by GTM workflow position. Search by name, trigger word, or vibe. Third-party plugin skills are deliberately excluded.

159 skills
EXECUTION touches client data, real sends, real systems
🌐
aicro-prospects-proposals-site
Clones the canonical proposal microsite and deploys to {slug}.aicro.co.
Build branded prospect-facing proposal microsites using the AICRO Signal-to-Scale template. NOT for one-off PDFs or static decks (use aicro-pptx). NOT for client-internal portals (use the client-portal Next.js app).
Show details
🚦
pre-enrichment-dnc
Looped search-api DNC + 90-day check before Clay credit burn.
Per-lead DNC + outreach-history check via the cortex search-api before Clay import. Saves enrichment credits by filtering blocked and recently-contacted leads first. Trigger words: pre-enrichment DNC, DNC check before Clay, pre-Clay check, filter blocked leads, dnc sweep.
Show details
📋
revbase-list-builder
Translates an ICP into Gigasheet filter configs against RevenueBase.
RETIRED: RevenueBase contract ended 2026-09; do not use. Build lists from the Lead Universe (knowledge.leads) instead. Build targeted B2B lead lists using RevenueBase data via Gigasheet.
Show details
🧪
clay-campaign-builder
Translates an ICP into a Clay table config and grading prompt chain.
Build AI grading prompt chains for Clay from campaign briefs. RevenueBase filter configs are retired (contract ended 2026-09). Saves deliverables to client campaign-briefs folder.
Show details
🛠️
campaign-builder
Build the campaign, explain the list, finish with every link and launch draft.
End-to-end campaign build pipeline. Chains 6 agents to go from raw client materials to outreach-ready sequences written to Google Sheets.
Show details
🚀
campaign-execution
End-to-end campaign run: list, qualify, enrich, copy, launch.
End-to-end execution of an outbound campaign covering list build, qualification, enrichment, copy generation, and platform launch. Claude handles 80% execution; GTME handles 20% strategy and review.
Show details
🔗
connector-pipeline
Grading, surfacing, normalization, copy for connector campaigns.
Run grading, variable surfacing, normalization, or copy generation for a connector campaign.
Show details
🎯
personalized-message-builder
Per-row reasoning grounded in the recipient's reality, not template merge fields.
AICRO's personalized outreach engine. Generates per-prospect, per-row email sequences (E1+E2+E3) and LinkedIn DMs (DM1+DM2) where every line is grounded in the specific recipient's reality.
Show details
📨
henry-message-builder
Per-row Henry AI E1+E2+E3 + DM sequences with locked discipline.
Generate Henry AI cold outreach sequences (E1+E2+E3 + LinkedIn DMs) for Henry AI campaigns. Per-row pipeline with locked discipline. One Claude Code subagent per row does reasoning + drafting + DMs in fresh context.
Show details
📢
campaign-launch-notification
Prepares launch copy in Linear and chat; Slack drafts only when requested.
Generate professional campaign launch notifications for client communications via Slack. Use when creating announcements for newly launched outbound marketing campaigns (email, LinkedIn, or multi-channel).
Show details
📞
phone-enrichment
Phone numbers for non-responding outbound leads from list data and Supabase.
Phone enrichment pipeline for non-responding outbound leads. Identifies leads who were contacted but didn't reply, cascades through data sources for phone numbers, and produces a call-ready CSV.
Show details
🛟
lead-remove-from-outreach
Pulls a lead from EmailBison, SmartLead, and HeyReach in one shot.
Remove a lead from active outreach across all sending platforms (EmailBison, SmartLead, HeyReach). Use when a lead replies on one channel and needs to be pulled from all others, or when a company/domain needs to be removed from active campaigns. Trigger words: remove from outreach, stop outreach, pull from campaigns, remove lead, stop sending, cross-platform removal, remove from all platforms, stop future emails, pause lead.
Show details
🚫
blocklist-lead
Per-client DNC in cortex plus cross-platform stop.
Add a lead to the per-client do-not-contact blocklist in cortex (knowledge.global_dnc via the upsert_blocklist_entry RPC) and stop all active outreach on EmailBison, SmartLead, and HeyReach. Use when a lead needs to be permanently blocked from future contact for a client. Trigger words: blocklist, add to blocklist, DNC, do not contact, add to DNC, block this lead, blacklist, block list, add to block list.
Show details
📡
campaign-performance-pull
Live workspace metrics from SmartLead, EmailBison, HeyReach by date.
Pull workspace-wide campaign performance metrics from SmartLead, EmailBison, and HeyReach APIs by date range.
Show details
🧭
airtable-navigator
Read-only navigator across AICRO's Airtable: clients, campaigns, replies, tasks.
Navigate AICRO's Airtable base for agency operations. Use when queries involve clients, campaigns, replies, inboxes, team members, tasks, client calls, or trends.
Show details
➕
emailbison-add-team-member
Adds a team member as admin to all EmailBison workspaces.
Add an AICRO team member as admin to all active EmailBison workspaces. Handles create-user on first workspace + invite/accept on remaining. Uses direct curl/Python — no MCP dependency.
Show details
➖
emailbison-remove-team-member
Removes a team member from all EmailBison workspaces.
Remove a team member from all active EmailBison workspaces. Resolves user ID from name/email, confirms workspaces, executes removal.
Show details
🎟️
slack-channel-invite
Bulk-invites team members to Slack channels by number.
Bulk-invite team members to Slack channels. Lists team from Airtable, fetches all Slack channels categorized and numbered, lets user pick by number, executes invites via Slack API.
Show details
🎭
playwright-cli
Browser automation for testing, form fills, scraping, screenshots.
Automates browser interactions for web testing, form filling, screenshots, and data extraction.
Show details
🔁
campaign-copy-sync
Edit copy once; sync the brief and the draft channels on purpose, never a started one.
Pull a client's edits from the campaign brief Google Doc into the local copy file, then explicitly sync that one copy revision to the branded brief (same Drive file) and the draft EmailBison / HeyReach specs. Records client approval or a Josh override per copy revision. Use when the operator says 'pull in the client's edits', 'sync the copy', 'push the copy to the brief', or 'record client approval'.
Show details
🎯
client-tam-pull
Empirically calibrated TAM pull for any client, multi-slice plus deduped.
RETIRED: RevenueBase contract ended 2026-09; do not use. Pull TAM from the Lead Universe (knowledge.leads) instead. Pull a client's full Total Addressable Market from RevenueBase via Snowflake. Use when the user says: pull TAM, build TAM, client TAM, total addressable market for {client}, TAM list, tam pull.
Show details
📅
cold-email-weekly-rhythm
The standing loop that keeps the fleet honest week over week.
The outbound operating cadence — Monday digest triage, Wednesday reply sweep, Friday 21-day campaign evals, biweekly rotation/renewal review. Use when: weekly rhythm, deliverability cadence, act on the digest, campaign 21-day eval, inbox rotation review.
Show details
🔬
deliverability-audit
The deliverability diagnostic, same math as the automated alerts.
Diagnose email infrastructure health for a client or the fleet — 1% rule, bounce bands, DNS auth, reputation tiers. Use when: reply rate dropped, bounces spiked, domain renewal decision, deliverability audit, is this domain healthy, inbox health check.
Show details
🚨
deliverability-incident-response
What to DO when deliverability breaks, in order, without panic.
Triage playbook when cold email breaks — reply collapse, bounce spike, blacklisted domain, blocked warmup. Use when: deliverability incident, landing in spam, bounces spiked, domain blacklisted, warmup blocked, reply rate collapsed.
Show details
📥
emailbison-add-leads
csv to emailbison campaign, with the mapping you confirmed
Upload a CSV of leads into an EmailBison campaign for a specific AICRO client. Use when the user says \"add leads to EmailBison\", \"upload leads to {campaign}\", \"push CSV to EmailBison\", or names a CSV + a campaign.
Show details
🚀
emailbison-campaign-deploy
One-shot deploy of a Step 5 spec into an EmailBison draft campaign.
Deploy draft EmailBison campaigns and assign sender inboxes from names such as Ward and Robert, using the client inbox inventory and verified API attachment. Also sets or copies one campaign's sending schedule (days, hours, timezone), rerunnable per campaign for bulk schedule changes. Use for campaign-builder output, approved campaign specs, or \"update the schedule on {campaign}\".
Show details
🥇
emailbison-reorder-priority
pause all, resume your picks first, everything else schedules behind
Yield EmailBison send priority to one or more chosen campaigns for an AICRO client, by pausing all active campaigns and resuming the chosen ones first so they take the front of the shared send queue. Use when the user says \"reorder priority\", \"yield priority to {campaign}\", \"make {campaign} send first\", \"prioritize this campaign in EmailBison\", or names a campaign that should jump the queue.
Show details
🔗
heyreach-add-leads
csv to heyreach campaign, with the mapping you confirmed
Bulk-upload leads to a HeyReach campaign for a specific AICRO client. Use when the user says \"add leads to heyreach\", \"upload to heyreach campaign\", \"push LinkedIn leads\", or names a CSV + a HR campaign.
Show details
🔗
heyreach-campaign-deploy
One-shot build of a LinkedIn campaign into a HeyReach DRAFT, standard sequence baked in.
Deterministic deploy of a HeyReach LinkedIn campaign from a JSON spec. Creates a new lead list named after the campaign, uploads the CSV leads, then creates the campaign in DRAFT with the chosen senders, schedule, exclusion flags, and the AICRO template sequence in one of two named shapes (standard, or event_note_only for conference campaigns). Verifies senders, rendered connection notes, exclusions and the sequence by readback, and resumes from its manifest without recreating anything. Always leaves the campaign in DRAFT, never starts it. Use when standing up a LinkedIn campaign in HeyReach from approved copy, building a HeyReach sequence, or pushing LinkedIn leads into a new campaign. NOT for EmailBison (use emailbison-campaign-deploy), NOT for adding leads to an existing HeyReach campaign (use heyreach-add-leads). Trigger words: heyreach deploy, deploy to heyreach, build heyreach campaign, push linkedin campaign, ship campaign to heyreach, new heyreach campaign, heyreach campaign deploy, create linkedin campaign.
Show details
🔌
heyreach-sender-reconnect
Pause a sender's campaigns, reconnect them, and get them back into every campaign.
Safely disconnect and reconnect a HeyReach LinkedIn sender account for any AICRO client. Pauses every running campaign the sender is in and saves a snapshot. After the operator reconnects the account by hand, it finds the new account id (a reconnect usually creates one), re-adds the sender to every paused campaign through the HeyReach public API, confirms the sender is attached, and resumes only those campaigns. Use when a sender needs their LinkedIn account reconnected, re-authed, or re-logged in HeyReach. Trigger words: reconnect sender, reconnect linkedin account, disconnect and reconnect, re-auth heyreach account, pause {name}'s campaigns, heyreach reconnect, sender reconnect, put {name} back in his campaigns.
Show details
📞
hubspot-segment-list-upload
sheet to hubspot segment, additive, with the mapping you confirmed
Upsert a lead list (Google Sheet or CSV) into a client's HubSpot, matched by email, LinkedIn URL and name, and optionally put every contact in one HubSpot List (segment). Use when the user says \"hubspot segment list upload\", \"hubspot segment upload\", \"upload to hubspot\", \"create a hubspot segment\", \"call down list to hubspot\", \"upsert these contacts into {client}'s hubspot\", or names a sheet + a client + a segment.
Show details
🔗
linkedin-network-export
client name in, every sender's rolodex out, overlap mapped
Pull the first-degree LinkedIn connections of a client's HeyReach sender accounts into one branded multi-tab Excel workbook, one tab per sender, with a cross-sender overlap analysis (who is connected to more than one sender).
Show details
🧭
revbase
List building and enrichment against RevenueBase's live API, through the gateway, never holding the key.
LIST BUILDING DISABLED: RevenueBase is switched off (contract ended 2026-09); count/preview/slice/lookalike/export return 503. The enrich wrappers still work through the remaining vendors. RevenueBase v2 list building and atomic enrichment via the cortex gateway -- count, preview, multi-slice union, lookalikes, async exports with auto-partitioning, and thin wrappers over the enrich atomic tools. Replaces the snowflake-revbase and client-tam-pull skills (Snowflake share retired). Use when the user says: revbase, revenuebase, pull a TAM, build a list, RevenueBase list, revbase slice, revbase export, revbase lookalike, revbase enrich.
Show details
❄️
snowflake-revbase global
SQL into the RevenueBase Snowflake share for ad-hoc pulls, list builds, and multi-slice exploration.
Direct SQL access to Josh's RevenueBase Snowflake share (account GITOZUX-HIB15754, service user ENRICH_SVC, role ENRICH_SVC_ROLE, key-pair auth). Live tables: REVENUEBASE_RELEASES_PERSONORGANISATION_DATA.PER.PER_LATEST (388M people rows, 59 cols) and ORG.ORG_LATEST (65M company rows, 61 cols), joined via p.RBID_ORG = o.RBID. Three modes: (1) ad-hoc querying for English-described slices, (2) single-slice list build with count-first protocol, (3) multi-slice filter exploration that runs N complementary filters in parallel, shows a counts comparison table, then pulls a deduped union. All exports emit a 73-col CSV format (Segment + 72 Gigasheet-standard columns) ready for /qualify, /surface, Clay. Read-only enforced at SQL-string layer in addition to the read-only role. Trigger words: snowflake, snowflake-revbase, revbase sql, revenuebase sql, sql revbase, query revbase, snowflake list, snowflake query, snowflake-revbase pull, revbase snowflake, multi-slice list, list build, slice.
Show details
GENERATION produces a deliverable for client or team review
📝
client-intake-assessment
Deep prospect research populated into a 6-section intake assessment .md.
Conduct deep research on a prospect company and generate a completed AICRO Client Intake Assessment as a markdown (.md) file. Output is an external-facing document formatted for client handoff or review.
Show details
🏆
deal-readiness-index
Pre-meeting M&A signal brief: 15 companies ranked by composite readiness.
Produce a pre-meeting signal intelligence brief (.docx) for M&A advisors, PE firms, VC firms, and corporate development teams in any sector thesis. Designed as a goodwill gift ahead of a first meeting — operator-to-operator tone, no pitch, no CTA.
Show details
💡
gtm-pilot-proposal
Pilot proposal .docx for prospects evaluating AICRO's Quarterly Growth Program.
Generate compelling pilot proposal documents for prospects evaluating AICRO's Quarterly Growth Program. NOT for existing paying clients (use gtm-client-strategy instead).
Show details
🎤
aicro-gtme-presentation
Executive GTME engagement deck: 11 core + 3 appendix + Thank You slides.
Generates executive-ready presentations for CEO/Founder approval of AICRO Go-To-Market Engagements.
Show details
🎓
client-onboarding-guide
23-slide onboarding deck: HeyReach, EmailBison, Reply Agent walkthrough.
Generate client onboarding presentations for AICRO's technology stack (HeyReach, Email Bison, AICRO Reply Agent). Designed from Chief Enablement Officer lens - self-service focused to minimize CS touchpoints.
Show details
🗺️
gtm-client-strategy
Quarterly execution strategy doc for paying clients with active campaigns.
Generate quarterly execution strategy documents for existing paying clients with active campaigns. NOT for prospects or new client onboarding (use gtm-pilot-proposal instead). Requires existing client data in Airtable.
Show details
🪪
icp-card-generator
Operator-grade ICP card a demand-gen exec can actually run Clay against.
Generate Demand Generation ICP Cards for AICRO clients and prospects. Adapts to any ICP type: investors, brokers, executives, SaaS buyers, technical personas, etc.
Show details
🎪
conference-list
Branded conference target .xlsx: 10-15 conferences scored 1-5 by ICP fit.
Generates a branded, scored conference target list (.xlsx) for a client or prospect based on their ICP. Use when a client or prospect needs a conference prioritization list as part of their GTM strategy.
Show details
📑
campaign-brief
External or internal campaign brief — branded .docx, Drive-uploaded.
Generate campaign briefs — external (client-facing Executive Overview) or internal (GTME execution brief). Produces branded .docx uploaded to Google Docs with markdown archive.
Show details
📚
deal-library
verified deal-library — per-deal markdown plus client tombstones plus consolidated Google Doc
Build a verified case-study / deal-library system for any AICRO client. Closes the proof-point fabrication gap by giving every campaign a [VERIFIED] source.
Show details
✉️
campaign-copywriting
3-email + 2-DM sequences with A/B variants for outbound campaigns.
Write cold email sequences and LinkedIn message sequences for outbound marketing campaigns. Use when the user needs copy, emails, LinkedIn messages, campaign content, or outreach sequences.
Show details
📧
cold-email-copywriter
Personalized cold emails from surfaced variables, GEX method, time cues.
Generate personalized cold emails from surfaced variables using YAML-driven ICP configs, time cue logic, and the GEX method. Pairs with /surface output and /copywrite pipeline.
Show details
💬
linkedin-copywriting
Cold LinkedIn DM sequences for HeyReach, blank connection requests, no-pitch first DM.
Generate cold LinkedIn outbound message sequences for HeyReach automation. Use for any LinkedIn outbound campaign, cold LI sequence, or HeyReach copy. NOT for email copy (use campaign-copywriting skill).
Show details
🧩
variant-strategy
personalization-token playbook mining client assets into reusable merge fields
Generate a client-facing Personalized Content Playbook (.xlsx) that mines the client's proprietary assets (named customers, partners, stats, deal narratives, pricing lines) into reusable merge-field data tokens for 1:N personalization at scale.
Show details
🤝
client-call-prep
Weekly call prep 1-pager: metrics, hot leads, recommendations, action items.
Single entry point for the weekly client call prep flow. Use when preparing for a client meeting or generating a weekly client summary — this is the canonical orchestrator.
Show details
📊
attribution-report
Joins your sends with the client's CRM and shows exactly what AICRO drove.
Generate the supplemental detailed attribution package for any AICRO client and date range. This skill does not generate MBRs or QBRs; use mbr-qbr-report-generator for those.
Show details
📈
mbr-qbr-report-generator
Cortex-native QBR workbook the client can audit, methodology included.
Generate the client-facing AICRO Quarterly or Monthly Business Review workbook: a branded Google Sheet with the shared Cortex funnel, sourcing cohorts, AICRO percent of client pipeline, revenue, and the attribution ledger.
Show details
🗂️
client-review-doc
navigation-hub doc linking every AICRO deliverable for a client
Generate a branded Client Review Document (.docx) that consolidates all AICRO deliverables, links, metrics, and context for a client into a single navigation hub.
Show details
🤖
tofu-agent-creator
Client-facing presentation explaining a TOFU identification agent.
Create client-facing presentations that explain TOFU identification agents. Use when you have an agent spec doc, discovery notes, or client brief and need to produce a shareable deck explaining how the agent works.
Show details
📅
client-strategy-notes
Auto-pulls Monday context the GTME team needs to walk into the week prepped.
Generate the weekly Client Strategy Notes DOCX — a branded per-client breakdown for the GTM team. Auto-pulls client meeting schedule, account owners, last-call context, weekly-brief + warm-reply links, and open-task record IDs from Airtable.
Show details
🧮
gtm-capacity-planning
Turns Airtable task backlog + team roster into the week's actual launch plan.
Generate a weekly GTM capacity plan for the AICRO team — produces a branded Account Coverage Plan .docx (owner per client, daily rhythm, priorities) + a Capacity Plan .xlsx with all open Airtable tasks, proposed weekly launches, brief readiness, and client scorecard.
Show details
💻
frontend-design
Production-grade frontend interfaces with high design quality.
Create distinctive, production-grade frontend interfaces with high design quality. Generates creative, polished code and UI design that avoids generic AI aesthetics.
Show details
✨
impeccable
distillation-grade frontend redesign with committed craft
Use when the user wants to design, redesign, shape, critique, audit, polish, clarify, distill, harden, optimize, adapt, animate, colorize, extract, or otherwise improve a frontend interface. Not for backend-only or non-UI tasks.
Show details
🎧
call-feedback
Reads every recorded call and tells the product team what customers actually said, with receipts.
Turn any client's recorded sales and customer calls into a themed product-feedback report for their product team, shareable with leadership.
Show details
📋
client-agenda
Last call plus the live board, out as a client pre-read the day before.
Build the Slack pre-read for a recurring client call, grounded in the last recorded call and the live board. Reads the prior Fireflies transcript and the Airtable commitment ledger, then emits an executive agenda under 200 words that goes out the day before. Use for a weekly or biweekly client call. Trigger words: client agenda, pre-call agenda, pre-read, agenda for tomorrow's client call, weekly client call prep agenda, send the agenda.
Show details
🧭
client-program-review
The account, campaign by campaign, with a launch queue the client can act on and an Airtable that matches Slack.
Periodic client performance review: what is working and what is not, campaign by campaign, a dated launch queue, and the decisions we need from the client. Produces a client-facing branded Google Doc plus an internal Slack post to the client's internal channel, and reconciles Slack commitments against Airtable Task Hub along the way. Use on a cadence (weekly, biweekly, monthly) or when taking over an account. NOT the MBR/QBR (use mbr-qbr-report-generator) and NOT the weekly brief (use client-call-prep).
Show details
🏢
duxre-weekly-report
The Duxre weekly, rebuilt from live systems with the judgement carried forward.
Build the Duxre weekly outbound workbook: activity, ICP scorecard, listing supply at broker and office level, qualified replies, and a client-shareable reconciliation of what the client portal shows against what the programme actually produced. Duxre only. NOT the canonical qualified-lead report (use qualified-lead-report) and NOT the MBR/QBR.
Show details
🚦
launch-queue
What went live, who raised a hand, and what is next in line, before every client call.
Build the weekly Launch Queue doc for a client call: what launched since the last call, positive replies, the campaigns built and waiting, decisions waiting on the client, what we are watching, and the status of the client's campaign ideas. Client-facing Google Doc, one per client per week, on command. Use before a weekly client call. Trigger words: launch queue, launch queue doc, the queue doc, what's launching, weekly client doc, prep the queue for {client}.
Show details
🚀
launch-slate
First 90 days of campaigns for a new client, sequenced against real sending capacity.
Propose a new client's first series of campaigns. Use at client launch or kickoff, NOT for quarter 2+ planning (use gtm-client-strategy) and NOT to sell the engagement (use gtm-pilot-proposal).
Show details
🖼️
linkedin-graphic
Branded LinkedIn stat / quote / link-share graphics, designed as React, exported as PNG.
Generate a branded AICRO LinkedIn graphic (PNG) from a plain-English request. Wraps the Remotion tool at scripts/linkedin-graphics/. NOT for cold copy (use linkedin-copywriting) or docs (use aicro-docx).
Show details
🟩
linkedin-inbox-audit
Proves what reached the system, names what did not, and hands back the leads nobody worked.
Audit every LinkedIn reply one sender account received, prove which ones reached the AICRO system, classify the rest, and ship a four-tab client-ready workbook. Answers the questions a client asks about their own LinkedIn inbox: how many replies came in, how many we captured, what the uncaptured ones actually are, and which leads the sales team has never seen. Use when a client asks why their LinkedIn messages are or are not in AICRO, when auditing coverage for a sender account, when quantifying inbound noise to justify a filtering decision, or when preparing a coverage readout for a client meeting. Trigger words: linkedin inbox audit, linkedin reply audit, linkedin coverage audit, are we capturing their linkedin, which linkedin replies are missing, linkedin messages not in AICRO, audit {name}'s linkedin, heyreach inbox audit, linkedin noise analysis, why aren't we piping in linkedin.
Show details
🗓️
meeting-agenda
Yesterday's recap plus today's board, out as a Slack-ready agenda.
Build a Slack-ready agenda for a call, grounded in the last recap and the live board. Asks who the audience is first, then carries forward open items and opens with what actually moved since yesterday.
Show details
📝
meeting-recap
Transcript in, Slack-ready recap out, executive overview plus action items by person.
Turn a meeting transcript or notes file into a Slack-ready recap: a 1-2 sentence executive overview plus action items broken down by person. Auto-detects internal team meetings (per-person owners) vs client calls (On You / On Us split).
Show details
📡
operator-brief
Internal status made visual, with its provenance showing.
Build an internal operator brief as a self-contained dated HTML file: launch boards, go-live decision briefs, status updates, audits, recaps. Use when the ask is a visual update for a named internal reader (Doug, Josh, Jacob, a client-facing operator) about the state of something we built. Fires on: status update, launch status, go-live brief, decision brief, visual update, what shipped, where are we on X, review the work across repos, exec summary, readout. NOT for client-facing or prospect-facing surfaces, which are aicro-brand rank 1.
Show details
🖼️
product-screen-capture
Live portal screens, rewritten into one fictional customer, gated so nothing real ships.
Capture high-resolution AICRO product screens from the live portal, rewrite every real name, company, message and figure into one fictional customer (Meridian), prove nothing real survived, and package a designer gallery plus zip. Use for website screens, deck or investor screenshots, designer handoffs, marketing images of the product. Trigger words: product screenshots, screens for the designer, website screens, capture the portal, scrubbed screenshots, example-data screens, screenshot the leads board/inbox/metrics. NOT for designing a page from references (use screenshot-to-code) and NOT for client-facing reports.
Show details
🧪
prototype-custom
Throwaway code that answers one design question, then dies.
Doug's customized variant of prototype, ends with delete-or-fold-into-code plus a NOTES.md capture (the plugin version instead commits the prototype to a throwaway branch as a primary source). Build a throwaway prototype to flesh out a design before committing to it. Routes between two branches, a single shareable HTML file for state/business-logic questions, or several radically different UI variations toggleable from one route. Use when the user wants to prototype, sanity-check a data model or state machine, mock up a UI, explore design options, or says \"prototype this\", \"let me play with it\", \"try a few designs\".
Show details
🎯
qualified-lead-report
Cortex-native qualified-lead report, one persistent sheet per client.
Generate the canonical AICRO Qualified Lead Report: a per-client Google Sheet ranking positive replies by genuine forward-interest intent, on the same Cortex funnel used by MBR/QBR.
Show details
🖼️
screenshot-to-code
References in, comparable renderings out, gated by the design stack.
Turn screenshots, mockups, Figma frames, or reference sites plus a written brief into comparable design renderings. Handles the hard case: harvesting different elements from several different sites and composing them into one coherent page, then rendering the directions side by side for a decision. Two modes, REBUILD (borrow structure, rebuild in AICRO tokens) and REPLICATE (pixel-faithful, only for surfaces we own). Prompt doctrine adapted from abi/screenshot-to-code (MIT). Use when the input to a design task is images rather than a written spec: reference sites, a screenshot of the current site, inspiration shots, a Figma export, a whiteboard photo, a hand sketch.
Show details
🧭
tam-expansion
Turn a validated universe into net-new lookalikes, precise from run one.
Expand a validated TAM by finding net-new lookalikes of confirmed companies. Use when the user wants to grow a TAM beyond what a keyword pull already found, run a lookalike expansion, brief Hermes, or turn a graded universe into net-new discovery.
Show details
ANALYSIS / UTILITY reads, synthesizes, or operational utility
🔎
client-data
Fetches client metrics from Airtable via webhook.
Fetches client campaign data from Airtable via webhook. Use when querying client information, campaign performance, reply rates, or any Airtable data for a specific client.
Show details
🪞
context-grounding
Verifies context and checks for existing solutions before building.
Verify context, check for existing solutions, and ground assumptions before building anything
Show details
who-built-this-before-me global
Check whether a project, tool, library, or product idea has already been built before the user invests time in it. Find existing solutions, extract standard patterns, benchmark the user's angle against prior art, and deliver a clear verdict (build / fork / contribute / use existing / investigate why it failed). Use this skill whenever the user describes something they're thinking of building, making, shipping, launching, or coding — especially phrases like "I want to build…", "I'm thinking of making…", "should I build X…", "is there already a…", "my idea is…", or when the user is scoping a new side project, startup, library, or tool. Trigger even if the user doesn't explicitly ask for prior art research — if they're about to build something, they should check first.
Show details
🧠
client-intelligence
Navigate client context, interpret metrics, synthesis patterns.
Navigate client context, interpret metrics, and understand synthesis patterns for AICRO client operations.
Show details
🔁
copy-variant-recall
Historical top-performing variants for client + ICP + sequence step.
Pull historical top-performing copy variants for a client + ICP + sequence step from prior AICRO campaigns. Returns subject lines, openers, send counts, and positive reply rates so the copywriter can scaffold new copy on what already worked.
Show details
✂️
stop-slop
Catches em-dashes, filler, and AI tells before external prose ships.
Post-generation copy QA. Runs banned-phrase + anti-pattern checks against any draft client-facing prose (cold copy, campaign briefs, ICP cards, weekly briefs, advisory docs, playbooks, decks).
Show details
⚖️
doctrine-audit
reads every rule against one client's actual files, no vibes-based compliance
Sweep one client folder against every always-active rule in .claude/rules/ and return a per-rule PASS/FAIL/PARTIAL/N/A table plus an owner-grouped follow-up task list.
Show details
humanizer global
Rewrite drafts to strip AI tells and add human voice. Use when the user provides a piece of writing (email, blog post, LinkedIn post, doc, campaign copy, message draft, etc.) and asks to humanize it, de-AI it, make it sound less like ChatGPT, fix the AI voice, or otherwise remove the telltale signs of LLM output. Also use when the user complains that something they or someone else wrote sounds like AI, robotic, generic, or hollow. Runs a two-pass edit (voice, then tells) plus a self-audit. Optionally matches the user's own writing style if a sample is provided. NOT for writing from scratch (use campaign-copywriting or linkedin-copywriting for outbound) and NOT for proofreading or grammar-only edits. Trigger words: humanize, de-ai, make it sound human, sounds like AI, sounds like ChatGPT, strip AI tells, AI voice, robotic, generic copy, hollow writing, fix the voice, less AI, rewrite this draft, edit for voice, kill the AI, soulless, fluffy, fix the tone, match my voice.
Show details
📐
prompt-engineering
Prompt design across model tiers: Standard, Clay, Grading, Single.
Design and structure prompts optimized for any model tier and platform. Handles four prompt types: Standard (system+input), Clay/Claygent (input only, web research), Grading (classification with JSON output), and Single (one-shot).
Show details
🎛️
design-taste
Design judgment plus anti-slop gate above the build skills.
The design waterfall, start to finish. Sets the Design Read and dials, resolves precedence (brand, accessibility, apple-design craft, taste, novelty), routes the build, then RUNS the gates rather than describing them: doctrine token parity, impeccable detect, computed WCAG contrast, stop-slop. Ends by offering the /architect-review pass. Use for any request to design, redesign, critique, or polish a visual surface, and invoke it directly as the entry point for design work.
Show details
🎨
aicro-brand
AICRO brand system reference: colors, typography, voice, design philosophy.
Applies AICRO's operator-built brand identity—colors, typography, voice, and design philosophy—to artifacts requiring the AICRO look-and-feel.
Show details
🖌️
ui-ux-pro-max
Design-system intelligence: palettes, typography, products, charts, UX rules.
Design-system intelligence for UI, web, and mobile work. Use when picking the visual system for a landing page, dashboard, client portal, prospect deck, marketing site, or mobile screen, BEFORE writing JSX/CSS.
Show details
🎬
design-motion-principles
motion-design audit through Kowalski / Krehel / Tompkins lenses
Expert motion and interaction design auditor based on Emil Kowalski, Jakub Krehel, and Jhey Tompkins' techniques. Use when reviewing UI animations, transitions, hover states, or any motion design work.
Show details
📄
aicro-docx
Global .docx pipeline: create, edit, read, upload, brand-apply Word documents.
Global .docx pipeline for AICRO — create, edit, read, and upload Word documents.
Show details
🖼️
aicro-pptx
Global .pptx pipeline: read, edit, create, brand-apply slide decks.
Use this skill any time a .pptx file is involved in any way — as input, output, or both. If a .pptx file needs to be opened, created, or touched, use this skill.
Show details
✏️
excalidraw-diagram
Excalidraw diagram JSON for workflows, architectures, concepts.
Create Excalidraw diagram JSON files that make visual arguments. Use when the user wants to visualize workflows, architectures, or concepts.
Show details
🧰
skill-creator
Create, modify, evaluate skills. Skill packaging and validation.
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, update or optimize an existing skill, package a skill for distribution, or validate skill structure.
Show details
🦸
using-superpowers
Disciplined skill usage: checks for applicable skills before responding.
Establishes disciplined skill usage across all interactions — not just computer-use tasks. NOT for simple factual questions, conversational follow-ups within active workflows, or meta-questions about how skills work.
Show details
💾
save-context
Save session state to session-context.md before /clear.
Save session state to session-context.md before /clear. Use when switching work streams, completing milestones, or before clearing context.
Show details
▶️
resume
Restore session context from session-context.md after /clear.
Restore session context from session-context.md after /clear. Displays all saved sections and asks which to pick up.
Show details
🍎
apple-design
Interaction physics and material craft, on AICRO ground.
Apple's interaction physics and material craft, translated to the web and resolved against AICRO's token set. Use when building or reviewing anything a user touches: transitions, hover and press states, drag or sheet interactions, depth and translucency, focus treatment, per-size typography, or how a surface reports stale, partial, and failed content. Rank 3 in the design-taste precedence order.
Show details
🏛️
architect-review
route by path, check against canon, verify against prod, post the verdict
Agentic architect PR reviewer. Built to hold the security-and-schema review coverage that leaves with Anderson.
Show details
🐞
bug-hunt
fan out finders, refute every claim, ship a severity-ranked fix plan
Cost-tiered, adversarially-verified codebase bug hunt. Fans out parallel reviewer agents by dimension, stress-tests every finding with independent skeptics, ranks survivors by severity (probability x impact), and emits a fixed remediation plan.
Show details
codex-usage
Show the user's live daily and weekly Codex quota remaining. Use when the user asks how much Codex usage is left. Do not use for session token accounting or historical activity.
Show details
🌐
cold-email-domains
Brandable .com variations, availability-confirmed against Porkbun.
Generate adjacent domain variations for a company's cold email sending infrastructure and confirm live availability via the Porkbun API. Use when a client or AICRO needs N new sending/forwarding domains (e.g. 'find 50 available domains for duxre.com', 'cold email domains', 'secondary domains', 'sending domains', 'domain variations'). Checks only; registration stays a human step in the Porkbun dashboard.
Show details
📊
daily-deck
the operator's morning loop, codified into one clickable view
One-click GTM operator daily-work deck. Codifies the manual /gtm-standup + /pipeline-check console loop into a single clickable dashboard with the metrics a terminal can't show: portfolio funnel, send/reply trend lines, per-client positive-reply-rate, qualified-lead counts, color-coded health, and alerts. Reads live from Cortex, renders a self-contained Chart.js HTML, and exports a dated snapshot into an Obsidian vault. Runs on Fable. Trigger words: daily deck, daily work deck, gtm dashboard, operator dashboard, daily standup dashboard, pipeline dashboard, one-click deck, morning deck, gtm daily view.
Show details
📐
diagram-design
Editorial explanatory diagrams as self-contained HTML plus inline SVG.
Create branded architecture, IT current-state, flowchart, sequence, state machine, ER/data model, timeline, swimlane, quadrant, radar/spider, polar chart (polar/radial lollipop), loop/flywheel, nested, tree, org chart, layer stack, Venn, pyramid/funnel, treemap, bar, line, Gantt and scatter charts, high-level, process, medallion, data flow, DP integration, DP security matrix, Sankey, fishbone, Wardley map, kanban, user journey, deployment, dependency graph, UML class, story map, or database schema diagrams as standalone HTML/SVG/PNG. Redraw .drawio/.drawio.png/.drawio.svg or Mermaid .mmd sources at a chosen size/detail; onboard brand tokens from a website; add semantic patterns, callouts, accessible motion, or sketchy/hand-drawn styling.
Show details
🛰️
fleet
See the whole engineering fleet by capability, with unknowns kept honest.
Roll up active panes, pull requests, local branches, and deployment ancestry by program across the configured repositories, and run a per-program conductor that is Doug's single point of contact for every leg: it reports on a regular cadence, delivers his answers to the legs, keeps PRs moving, and closes finished workspaces.
Show details
🛡️
outreach-blocklist-check
One endpoint answers "blocked?" and "already contacted?" per client, fail-safe.
Check whether a lead is blocklisted (DNC) or already contacted for a client via the Cortex search-api edge function. The canonical go/no-go check path: read-only, per-client, fail-safe. Use before any send, upload, or when asked 'is this lead blocked?' or 'have we contacted them?
Show details
📦
pick-ui-library
The package pick, so nobody hand-rolls a toast again.
Pick the package for a frontend task from a curated list: primitives, toasts, command menus, charts, virtualization, drag and drop, OTP inputs, state, styling, theming. Use when asked what to install or reach for, or when about to hand-roll a component that a trusted library already solves. Picks the PACKAGE. ui-ux-pro-max picks the visual system, dataviz designs the chart.
Show details
🔮
pre-mortem
convert imagined failure into owned, instrumented, resourced action
Run a production-grade pre-mortem on any initiative: product launch, GTM motion, hire, architecture choice, M&A move, market entry, pricing change, or strategy pivot. Composable: callable as a sub-step from any strategy or GTM skill.
Show details
🩺
qa-campaign
Python gathers and decides readiness. A subagent writes the judgment. No model API key, ever.
AICRO QA agent for outbound campaigns. Three input modes: Airtable task, manual brief Doc URL plus campaign platform URLs, or a local run manifest (Mode C) whose deterministic checks decide email and LinkedIn readiness separately. Triggered by an Airtable button via n8n → GitHub Actions, or run locally.
Show details
🚉
staging-vs-main
Who built what, what it does, and whether it can ride the next promote.
Daily view of what is on client-portal staging but not on main (so not in production), grouped by feature (Linear project), with each ticket's PR author, a plain-English line per PR, its engine (gtme-cortex) counterpart PRs and whether they are live, and a recommendation per ticket plus an overall promote call. Use when Doug says "what's on staging", "staging vs main", "what's waiting to promote", "can we promote", "release check", "/staging-vs-main". NOT for what already shipped (pr-recap) or client-facing release notes (client-portal release-notes).
Show details
📏
statistical-power
How many before a difference is real, and the smallest one this sample can see.
Sample-size and statistical power calculations for planning studies. Use whenever someone asks "how many subjects/samples/replicates do I need", wants an a priori power analysis, a minimum detectable effect (MDE), a power curve, or needs to justify a sample size for a grant, IRB protocol, or pre-registration. Covers closed-form power for t-tests, ANOVA, proportions, correlations, chi-square, and regression, plus simulation-based (Monte Carlo) power for designs with no formula, logistic/Poisson regression, mixed models, cluster-randomized trials, survival, and interactions. Use this skill even when the request only mentions an effect size, alpha, or "80% power" without saying "power analysis" explicitly. For laying out the study (randomization, blocking, factorial/DOE, crossover, sequential designs) use experimental-design; for analyzing data already collected and reporting it use statistical-analysis.
Show details
🧪
validate-outcomes
Fable judges each outcome statement against live warm replies + closed-won, stamps validated or flags with cited data.
Validate a client's per-persona OUTCOME STATEMENTS in playbook/outcome-translation.md against live data, exactly as .claude/doctrine/copy/outcome-level-value-prop.md mandates.
Show details
archify global
Create polished, validated architecture, workflow, sequence, data-flow, and lifecycle/state diagrams as explorable standalone HTML with inline SVG, dark/light themes, optional trace motion, and PNG/JPEG/WebP/SVG/WebM export. Accept plain-language requirements or pasted Mermaid flowchart, sequenceDiagram, and stateDiagram input; inspect repository evidence when the diagram must reflect real code. Use when the user asks to visualize system architecture, infrastructure, cloud/security/network topology, technical workflows, API call sequences, request lifecycles, data pipelines, ETL/ELT, data lineage, state machines, or to convert/beautify Mermaid.
Show details
batch-grill-with-docs-codex global
Batch plan hardening against project domain docs, followed by an adversarial Codex review loop. Use when the user asks for batch grilling, frontier-round questions, or the batch grill-with-docs Codex workflow. Skip for code review and trivial changes.
Show details
build-tam global
Build a Total Addressable Market list by sourcing accounts and contacts from providers like Crustdata, Dropleads, and PDL.
Show details
caveman global
> Ultra-compressed communication mode. Cuts token usage ~75% by dropping filler, articles, and pleasantries while keeping full technical accuracy. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief", or invokes /caveman.
Show details
clay-to-deepline global
Convert a Clay table configuration into local Deepline scripts. Handles extraction (MCP or script), documentation, action mapping, script generation, and parity validation against Clay ground truth.
Show details
codex-build global
Delegate implementation of a frozen specification to a separate Codex session, then review and iterate on its diff. Use for substantial work orders, mechanical migrations, known-reproduction bug fixes, or when the user explicitly asks Codex to build a locked plan. Skip tiny edits and unresolved design work.
Show details
codex-review global
Run an adversarial cross-model review loop over an implementation plan before code is written. Use when the user already has a plan, requests a Codex plan review, or wants a second-model sanity check for high-stakes work. Skip reviews of already-written code and trivial changes.
Show details
conductor global
Conduct a fleet of parallel Claude sessions in Herdr — write session handoffs, route cross-session messages, grant merge-freeze windows, resolve claim/number collisions. Use when acting as a conductor over other sessions, when a handoff file names peer sessions, or when a peer session requests coordination (freeze, unhold, ownership). For raw Herdr CLI mechanics, invoke the herdr skill.
Show details
deepline-ads-audiences global
Use this skill when building, enriching, auditing, or uploading B2B paid ads audiences to Google Customer Match, Meta/Facebook Custom Audiences, or LinkedIn Matched Audiences. Triggers on phrases like '/deepline-ads-audience', '/deepline-ads-audiences', 'upload this audience', 'create custom audiences', 'personal email hashes', 'increase Facebook match rate', 'Google ads audience', 'Meta audience', 'FB audience', or any workflow that turns CRM/customer/contact data into paid ads upload lists. Skip for outbound prospecting sequences, cold email, or pure campaign copywriting.
Show details
deepline-analytics global
Use this skill when answering business analytics, RevOps, GTM metric, pipeline, revenue, funnel, customer, or warehouse questions with Deepline. Triggers on phrases like 'query Snowflake', 'analyze pipeline', 'total ACV', 'break down by quarter', 'use the semantic layer', 'run a semantic query', or any use of snowflake_get_semantic_layer / snowflake_run_semantic_query. Skip prospecting, enrichment, contact finding, outbound, or personalization workflows; use deepline-gtm for those.
Show details
deepline-feedback global
Send feedback or bug reports to the Deepline team, including session transcript and environment info. Use when the user asks to report a bug, send product feedback, or share the current Claude/Cowork session with Deepline support.
Show details
deepline-gtm global
Use for GTM prospecting, enrichment, qualification, account/contact research, waterfalls, email/LinkedIn lookup, personalization, scoring, campaigns, CSVs, and Deepline plays/scripts. Source discovery: deepline-pre-research. Providers: adyntel, ai_ark, allegrow, apify, attio, aviato, bettercontact, bloomberry, bluesky, browserbase, builtwith, cloudflare, contactout, crustdata, crustdata-v2, crustdata-v3, customer_db, dataforseo, datagma, deepline_native, deeplineagent, discolike, dropleads, emailbison, enformion, exa, findymail, firecrawl, forager, fullenrich, generic_http, google_ads_audiences, hackernews, heyreach, hubspot, hunter, icypeas, instantly, ipqs, leadmagic, lemlist, limadata, linkedin_ads_audiences, linkedin_scraper, lusha, meta_audiences, openmart, opensosdata, openwebninja, parallel, peopledatalabs, predictleads, prospeo, rocketreach, salesforce, scrapecreators, sentrion, serper, slack, smartlead, snowflake, sumble, theirstack, trestle, twitterapi, upcell, wiza, wizleads, zerobounce.
Show details
deepline-monitors global
ACCESS-GATED beta. Deepline Monitors (dashboard name: Signal Radars) — provider event feeds (job posts, email replies, funding, intent) that stream into your warehouse and trigger plays. Only use if you have monitor access: run `deepline monitors status` first; if it reports no access, do NOT use this recipe — every command returns monitor_access_required. Ask a Deepline admin (Admin → Rollouts) for access.
Show details
deepline-plays global
Create custom Deepline plays/scripts that combine multiple tools and/or other plays, with durable datasets, fallback logic, joins, projections, and custom run/export behavior.
Show details
deepline-pre-research global
Use when the user wants a last30days-style pre-research pass in Deepline: discover the critical public, private, CRM, workflow, social, and web data sources for a research/enrichment job; compare provider coverage; estimate Deepline credit cost; recommend the source plan before building or running the workflow; or build custom language/messaging from buyer, competitor, community, and CRM evidence. Triggers: pre-research, source discovery, provider strategy, research data sources, ScrapeCreators, X/Twitter data, Reddit comments, public and private datasets, CRM data, workflow data, custom language, messaging language, pain language.
Show details
deepline-quickstart global
Run a quick Deepline demo recipe to show the user how Deepline works.
Show details
diagnosing-bugs global
Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.
Show details
domain-modeling global
Build and sharpen a project's domain model. Use when the user wants to pin down domain terminology or a ubiquitous language, record an architectural decision, or when another skill needs to maintain the domain model.
Show details
eod-slack-update global
Research everything Josh did over a time window — git commits, GitHub PRs, Claude Code sessions, Herdr panes, Linear issues — and write a tagged, actionable status update, then DM it to him in Slack via the cortex bot. Use when he says /eod-slack-update, 'EOD update', 'end of day update', 'what did I do today', 'write my standup', 'weekly update', or asks for a team update covering the last N hours/days.
Show details
find-qualified-titles global
Use when finding real role-holders at known company domains from an ICP, especially prompts like 'find all job titles at these companies', 'find qualified titles', 'find RevOps or marketing-ops buyers', or when exact title discovery should precede paid people search.
Show details
find-skills global
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.
Show details
grill-me global
A relentless interview to sharpen a plan or design.
Show details
grill-me-codex global
Interview the user until a plan is fully specified, then run an adversarial cross-model Codex review loop. Use when the user asks to be grilled and then have Codex review the plan, especially for high-stakes designs. Skip already-written code review and trivial changes.
Show details
grill-with-docs global
A relentless interview to sharpen a plan or design, which also creates docs (ADR's and glossary) as we go.
Show details
grill-with-docs-codex global
Interview the user against the project's domain model and living docs, update those docs as decisions settle, then run an adversarial cross-model Codex review loop. Use for high-stakes planning with established terminology or ADRs. Skip already-written code review and trivial changes.
Show details
grilling global
Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.
Show details
handoff global
Compact the current conversation into a handoff document that another agent can use to continue the work. Use when the user requests a handoff, conversation compaction, or preparation for a fresh agent session.
Show details
herdr global
Control Herdr, a terminal multiplexer for coding agents. Use only when the user explicitly mentions Herdr or asks to use Herdr to inspect or control panes, tabs, workspaces, commands, or another agent. Do not use merely because a task could benefit from a background terminal, delegation, or parallel work. Requires HERDR_ENV=1.
Show details
improve-codebase-architecture global
Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.
Show details
linkedin-url-lookup global
Resolve LinkedIn profile URLs from name + company with strict identity validation to avoid false positives.
Show details
n8n-code-javascript global
Write JavaScript code in n8n Code nodes. Use when writing JavaScript in n8n, using $input/$json/$node syntax, making HTTP requests with $helpers, working with dates using DateTime, troubleshooting Code node errors, or choosing between Code node modes.
Show details
n8n-code-python global
Write Python code in n8n Code nodes. Use when writing Python in n8n, using _input/_json/_node syntax, working with standard library, or need to understand Python limitations in n8n Code nodes.
Show details
n8n-expression-syntax global
Validate n8n expression syntax and fix common errors. Use when writing n8n expressions, using {{}} syntax, accessing $json/$node variables, troubleshooting expression errors, or working with webhook data in workflows.
Show details
n8n-mcp-tools-expert global
Expert guide for using n8n-mcp MCP tools effectively. Use when searching for nodes, validating configurations, accessing templates, managing workflows, or using any n8n-mcp tool. Provides tool selection guidance, parameter formats, and common patterns.
Show details
n8n-node-configuration global
Operation-aware node configuration guidance. Use when configuring nodes, understanding property dependencies, determining required fields, choosing between get_node detail levels, or learning common configuration patterns by node type.
Show details
n8n-validation-expert global
Interpret validation errors and guide fixing them. Use when encountering validation errors, validation warnings, false positives, operator structure issues, or need help understanding validation results. Also use when asking about validation profiles, error types, or the validation loop process.
Show details
n8n-workflow-patterns global
Proven workflow architectural patterns from real n8n workflows. Use when building new workflows, designing workflow structure, choosing workflow patterns, planning workflow architecture, or asking about webhook processing, HTTP API integration, database operations, AI agent workflows, or scheduled tasks.
Show details
niche-signal-discovery global
Discover niche first-party signals that differentiate Closed Won vs Closed Lost accounts for ICP analysis. Use when the user provides won/lost customer domain lists and wants differential signals (website content, job listings, tech stack, maturity markers) to build account scoring models and prospecting criteria. Triggers: ICP analysis, niche signals, won vs lost analysis, differential signals, signal discovery, ICP signal report, account scoring signals, lead scoring, first-party signals, buyer signals. Before reading this file, first read deepline-gtm to understand the Deepline CLI tool and how to use it. Then read this file for guidance on the task.
Show details
portfolio-prospecting global
Find companies backed by a specific investor or accelerator, then find contacts and build personalized outbound.
Show details
prototype global
Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.
Show details
research global
Investigate a question against high-trust primary sources and capture the findings as a Markdown file in the repo. Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent.
Show details
setup-matt-pocock-skills global
Configure AGENTS.md or CLAUDE.md and docs/agents so engineering skills know the repository's issue tracker, triage vocabulary, and domain-doc layout. Explicitly invoke before first use of the associated engineering workflows or when their project context is missing.
Show details
supabase global
Use when doing ANY task involving Supabase. Triggers: Supabase products (Database, Auth, Edge Functions, Realtime, Storage, Vectors, Cron, Queues); client libraries and SSR integrations (supabase-js, @supabase/ssr) in Next.js, React, SvelteKit, Astro, Remix; auth issues (login, logout, sessions, JWT, cookies, getSession, getUser, getClaims, RLS); Supabase CLI or MCP server; schema changes, migrations, declarative schemas, security audits, Postgres extensions (pg_graphql, pg_cron, pg_vector); debugging and troubleshooting errors or unexpected behavior on Supabase projects (HTTP errors, Postgres errors, RLS surprises, permission denied, schema cache issues, timeouts, Edge Function crashes, Realtime drops, Storage failures) and reading or querying logs (Logs Explorer, ClickHouse).
Show details
supabase-postgres-best-practices global
Postgres best practices maintained by Supabase, for Postgres running anywhere. Load this skill BEFORE writing or changing anything that lives in a Postgres database: creating or altering tables and columns (including choosing column types), schema design, migrations and declarative schema files, RLS policies and the tests that verify them, indexes, triggers, database functions, queues and scheduled jobs (pg_cron, pgmq), vector/semantic search (pgvector), and restoring dumps (pg_restore) or importing data. Also load it when diagnosing slow queries, high CPU, timeouts, EXPLAIN plans, connection exhaustion, locking, bloat, or rows visible to the wrong user or tenant. This is not just a performance guide — schema, migration, security, and SQL authoring tasks need these rules too, even for a one-column change or a single query.
Show details
tdd global
Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.
Show details
teach global
Teach the user a skill or concept as a stateful, multi-session learning workspace. Explicitly invoke when the user asks to learn a topic through an ongoing curriculum with durable progress artifacts.
Show details
to-spec global
Turn the current conversation into a specification and publish it to the configured issue tracker without a new requirements interview. Explicitly invoke when the user asks to synthesize the existing discussion into a spec or PRD.
Show details
to-tickets global
Break a plan, specification, or conversation into tracer-bullet tickets with explicit blocking edges and publish them to the configured tracker. Explicitly invoke when the user asks to create implementation tickets or decompose work for parallel execution.
Show details
triage global
Move issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.
Show details
wayfinder global
Plan work larger than one agent session as a shared map of decision tickets, then resolve those tickets until the route to the destination is clear. Explicitly invoke for large, ambiguous efforts that need decision mapping before implementation.
Show details
zoom-out global
Give a broader architectural map of unfamiliar code using the project's domain vocabulary, including relevant modules and callers. Explicitly invoke when local implementation details need to be placed in their larger system context.
Show details
No skills match that search.