cue-ai 0.9.0 → 0.9.2

package/resources/skills/skills/eu-funding/hu-grant-finder/SKILL.md

@@ -0,0 +1,114 @@
+---
+name: hu-grant-finder
+description: 'Use when user says "find grants", "pályázat", "támogatás", "GINOP", "DIMOP", "Hiventures", "EIC", or what funding a HU/SK company can apply for. Maps grants, loans, VC, not tenders.'
+tags: [eu-funding, grants, palyazat, ginop, hungary, venture-capital]
+---
+
+# Hungarian / EU grant & funding finder
+
+Map the funding a company can actually get, by channel and by stage. Built for
+agentic-AI / IT companies in Hungary (extendable to SK). **Grants are not tenders**, 
+for public procurement (selling to the state) use [[ted-tender-search]].
+
+## First: pick the right channel (they get confused constantly)
+
+| Channel | What it is | You... | Repay? |
+|---|---|---|---|
+| **Közbeszerzés / tender** | TED procurement | *sell* services to the state | no (you earn) |
+| **Támogatás / grant** | GINOP/DIMOP, EU grants | get non-repayable money for a project | no |
+| **Kedvezményes hitel** | MFB / GINOP loan programs | borrow cheap (often 0%) | **yes** |
+| **Kockázati tőke / VC** | Hiventures, EIC, EU funds | take equity investment | no, but give **equity** |
+
+## The funding ladder (Hungarian micro/IT company)
+
+```
+ 0 lezárt év          1+ lezárt üzleti év            deep-tech / scalable product
+ ──────────           ──────────────────             ───────────────────────────
+ Vállalkozóvá válás   KKV grants & 0% loans          Hiventures (GINOP 2.5.1, ≤€300k)
+ (NFSZ, ~4,9M Ft,     (DIMOP/GINOP — when a round    EIC Accelerator (≤€2,5M + equity)
+  if jobseeker)        is open; check forrás left)   → needs TRL6 product + company
+ + "saját honlap"      EV closed year COUNTS;         → "only-software" firms may be
+   voucher (0,4–2M)    don't found a fresh Kft and     excluded from VC — verify
+                       reset the closed-year clock
+```
+
+## Canonical sources (check live, frames run "until funds run out")
+
+- **palyazat.gov.hu**, official Széchenyi Terv Plusz portal; the *active list* is the
+  source of truth for open/closed. JS-rendered: status reliable, deep params need the PDF.
+- **nkfih.gov.hu**, innovation & VC calls (GINOP 2.x). Often slow/times out, be patient.
+- **mfb.hu**, the loan programs (GINOP 1.4.x KKV Technológia Hitel).
+- **hiventures.hu**, the early-stage VC fund manager (runs GINOP 2.5.1, ≤€300k).
+- **eic.ec.europa.eu**, EIC Accelerator (deep-tech, ≤€2,5M grant + equity).
+- **ec.europa.eu/info/funding-tenders**, EU Funding & Tenders Portal.
+- Aggregators (palyazatmenedzser, palyaz.hu, eumanagement) are useful for the *active
+  list* but are **paywalled / conflicting** on detail and status, confirm against official.
+
+## Slovakia (SK) sources
+
+The same channels, ladder, and gotchas apply; swap the Hungarian portals for the SK ones:
+
+| Channel | HU | SK equivalent |
+|---|---|---|
+| EU-funds portal (source of truth) | palyazat.gov.hu | **ITMS21+** (`portal.itms21.sk`, `itms2014.sk`) |
+| Operational programme | Széchenyi Terv Plusz | **Program Slovensko 2021-2027** (`eurofondy.gov.sk`) |
+| Cheap loan / digital | MFB, GINOP 1.4.x | **Plán obnovy / VAIA** via SIH (digital, rate cut + up to 30% loan forgiveness) |
+| Early-stage VC | Hiventures | **Venture to Future Fund** (`vff.sk`) via Slovak Investment Holding (EIB + MoF, IT focus) |
+| Micro / SME | KKV grants | **Slovak Business Agency (SBA)** microloans €2,500-50,000; **National Holding Fund** €20k-1.5M (IT) |
+| Innovation / R&D advisory | NKFIH | **SIEA** regional advisory; **APVV / VAIA** R&D grants |
+
+EU-level (EIC, Digital Europe, Horizon) is shared across both countries. Confirm live
+status on the SK portal, the same "funds run out" caveat applies.
+
+## Hard-won gotchas (verified this domain, they bite every time)
+
+- **Closed-year wall.** Most KKV grants/loans (DIMOP 1.2.6, GINOP 1.4.3) require ≥1
+  full closed business year + ≥1 employee. A brand-new company cannot apply day one.
+- **EV counts.** An egyéni vállalkozó's closed tax year *is* a closed business year, and
+  EV is usually an eligible form. Don't throw that history away by founding a fresh Kft.
+- **Region split.** For DIMOP/GINOP the line is Budapest (`/C`) vs. every other county
+  (`/B`, "less developed"), wealthy counties like Győr-Moson-Sopron are in the *favoured*
+  `/B`, but ~65% of some loan frames is ring-fenced for the 4 most-disadvantaged regions.
+- **Timing kills.** Frames close on *forráskimerülés* (funds exhausted), often before the
+  stated deadline. The spring-2026 round closed DIMOP 1.2.6/B, GINOP 1.2.4, 2.1.3, 2.1.4.
+  Always re-check the active list the same day.
+- **VC ≠ free.** Hiventures/EIC take equity and need a TRL6 working product + a company.
+- **"Only software" exclusion.** GINOP 2.5.1 (Hiventures) excludes firms developing *only*
+  immaterial assets (software). A pure SaaS/agent-wrapper play may not qualify, needs a
+  data/hardware/vertical asset beyond code. Verify the exact clause in the call PDF.
+- **Deep-tech ≠ SaaS for EIC.** An LLM-wrapper is SaaS, not deep tech; EIC wants
+  breakthrough + technological risk. Hungary is a "widening" country → EIC **Pre-Accelerator**
+  is the realistic stepping stone before the full Accelerator.
+
+## Research procedure
+
+1. **Classify the ask**, tender / grant / loan / VC (table above). Wrong channel wastes hours.
+2. **Place on the ladder**, how many closed business years? That gates everything.
+3. **Pull the active list**, palyazat.gov.hu (grants/loans) + hiventures.hu / eic (VC).
+   Filter to the company's region and form.
+4. **Confirm status + eligibility** from the official call PDF (not aggregators): open?
+   closed-year requirement? region? amount? intensity? exclusions?
+5. **Rank by fit**, run `/roi-estimator` when listing 3+ options.
+6. **Tag confidence**, amounts/deadlines move weekly; mark unverified figures and name the
+   one thing to confirm by phone (NKFIH / kormányhivatal) before the user acts.
+
+## Rules
+
+- Never quote a grant amount, deadline, or eligibility from memory, it is stale. Fetch live
+  and tag confidence; a "verify at palyazat.gov.hu" beats a confident wrong number.
+- State the channel explicitly every time (grant vs loan vs equity), the user conflates them.
+- Match the portal to the company's country: HU goes to palyazat.gov.hu, SK to ITMS21+ /
+  eurofondy.gov.sk. Don't quote a Hungarian program to a Slovak company or vice versa.
+- Surface the closed-year and region gates *before* detailing amounts; they disqualify fastest.
+- For procurement/tenders, hand off to [[ted-tender-search]]. For deep research sweeps, [[defuddle]].
+
+## Example
+
+User: "I just started a Hungarian IT company. What grants can I apply for?"
+
+Classify (grant, not tender), place on the ladder: 0 closed years means the big KKV
+grants (DIMOP, GINOP) are gated out, so day-one is vállalkozóvá válás (if a jobseeker
+founder) plus the saját-honlap voucher; the real money opens after one closed business
+year. If the company is a scalable AI *product* at TRL6, route to Hiventures (GINOP
+2.5.1, equity), watching the "only software" exclusion. Pull the live active list from
+palyazat.gov.hu, confirm eligibility in the call PDF, and tag every amount's confidence.

package/resources/skills/skills/eu-funding/hu-grant-finder/evals.md

@@ -0,0 +1,26 @@
+# Activation evals — hu-grant-finder
+
+Run by judging each query against the `description:` only: does it fire?
+Pass bar: >=90% of TRIGGER fire, <=10% of NO-TRIGGER false-fire.
+
+## TRIGGER (must fire)
+
+- what grants can my Hungarian company apply for?
+- pályázat IT cégnek
+- GINOP DIMOP support amounts
+- is there funding for a new startup in Hungary?
+- Hiventures venture capital eligibility
+- EIC Accelerator for my AI company
+- támogatás induló vállalkozásnak
+- what EU funding can a Slovak company get?
+
+## NO-TRIGGER (must NOT fire — route elsewhere)
+
+- find EU tenders to bid on              → ted-tender-search
+- search TED notices                     → ted-tender-search
+- write a grant application PDF           → gstack/document-generate
+- scrape palyazat.gov.hu                  → gstack/scrape
+- review my code                         → none
+
+Last run (after EIC re-added, boundary = "not tenders"):
+8/8 trigger, 0/5 false-fire = 100%.

package/resources/skills/skills/eu-funding/ted-tender-search/SKILL.md

@@ -0,0 +1,80 @@
+---
+name: ted-tender-search
+description: 'Use when user says "find EU tenders", "TED search", "public procurement", "közbeszerzés", or wants EU procurement notices by CPV, country, or keyword. Queries the free official TED API (no key).'
+tags: [eu-funding, tenders, procurement, ted, hungary]
+---
+
+# TED tender search
+
+Search live EU public-procurement notices through the **official TED API** (Tenders
+Electronic Daily). Free, no API key, no scraper. This is the canonical source, 
+Apify/Firecrawl and most "TED MCPs" just resell this same data.
+
+- **Endpoint:** `POST https://api.ted.europa.eu/v3/notices/search`
+- **Auth:** none
+- **Notice page:** `https://ted.europa.eu/en/notice/<publication-number>` (e.g. `568232-2025`)
+
+## Run a search (one command)
+
+`scripts/ted-search.sh` builds the query, calls the API, and prints the newest 15
+notices ranked, with buyer + notice URL. Args: buyer country, CPV codes, scope, limit.
+
+```bash
+scripts/ted-search.sh HUN "72000000 48000000 73000000" ACTIVE 50
+```
+
+It prints `TOTAL match: <n>` (the full count, not just the page). For custom filters
+(`place-of-performance`, `FT ~ (...)`) build the body by hand using the fields below;
+the request is `POST https://api.ted.europa.eu/v3/notices/search` with JSON
+`{query, fields, limit, scope}`, `scope` = `ACTIVE` (open) or `ALL` (historical).
+
+## Query language (verified fields)
+
+| Filter | Field | Example |
+|---|---|---|
+| CPV code | `classification-cpv IN (...)` | `classification-cpv IN (72000000 48000000)` |
+| Buyer's country | `organisation-country-buyer IN (...)` | `organisation-country-buyer IN (HUN)` |
+| Where work is performed | `place-of-performance IN (...)` | `place-of-performance IN (HUN SVK)` |
+| Full text (title/summary) | `FT ~ ("..." "...")` | `FT ~ ("artificial intelligence" "mesterséges intelligencia")` |
+
+- Combine with `AND` / `OR`. Country codes are **ISO 3-letter** (`HUN`, `SVK`, `AUT`).
+- `organisation-country-buyer` (who buys) is usually a sharper filter than
+  `place-of-performance` (where work happens), the latter also catches EU-institution
+  contracts that merely touch the country.
+- `FT ~ (...)` searches a narrow title/summary field, expect few hits; it is for
+  pinpointing, not broad discovery.
+
+## CPV codes for IT / AI / software work
+
+| CPV | Domain |
+|---|---|
+| `72000000` | IT services: consulting, software development, internet, support |
+| `48000000` | Software packages and information systems |
+| `73000000` | R&D services |
+
+## Ranking
+
+The API has no relevance sort, so `ted-search.sh` ranks by recency (the publication
+number embeds the year, `NNNNNN-YYYY`) and picks the `eng` then `hun` value from the
+multilingual title/buyer objects (`{"eng":[...], "hun":[...]}`). Edit the script's
+final `python3` filter to sort by a different field.
+
+## Rules
+
+- Always hit the official API first; do not reach for a paid scraper or a TED MCP for
+  search, they wrap this same free endpoint.
+- Use `scope: "ACTIVE"` for biddable tenders; `"ALL"` only to size history.
+- Report `totalNoticeCount`, not just the page you pulled, so the user sees the real universe.
+- Every tender you surface carries its publication number + notice URL. No number, not in the list.
+- A new company rarely wins a tender directly (references/turnover needed), flag the
+  subcontractor route. For *grants/funding* (not procurement), use [[hu-grant-finder]].
+
+## Example
+
+User: "Find open IT tenders from Hungarian buyers."
+
+Query `classification-cpv IN (72000000 48000000 73000000) AND organisation-country-buyer IN (HUN)`
+with `scope: "ACTIVE"`, report `totalNoticeCount`, then list the newest 15 with
+publication number + notice URL. Add `place-of-performance IN (HUN SVK)` to widen to
+Slovakia, or `FT ~ ("artificial intelligence" "mesterséges intelligencia")` to pinpoint
+AI notices.

package/resources/skills/skills/eu-funding/ted-tender-search/evals.md

@@ -0,0 +1,26 @@
+# Activation evals — ted-tender-search
+
+Run by judging each query against the `description:` only: does it fire?
+Pass bar: >=90% of TRIGGER fire, <=10% of NO-TRIGGER false-fire.
+
+## TRIGGER (must fire)
+
+- find EU tenders for IT companies
+- search TED for Hungarian procurement
+- közbeszerzés keresés szoftverre
+- what public procurement notices are open in Slovakia?
+- find tenders by CPV code 72000000
+- are there open EU contracts my company can bid on?
+- TED search for AI tenders
+- list active procurement notices from Hungarian buyers
+
+## NO-TRIGGER (must NOT fire — route elsewhere)
+
+- find grants for my startup            → hu-grant-finder
+- what GINOP can I apply for?           → hu-grant-finder
+- find me a venture capital investor    → hu-grant-finder
+- write me a bid proposal               → content/article-writer
+- scrape this website                   → gstack/scrape
+- how do I register a Kft?              → none
+
+Last run: 8/8 trigger, 0/6 false-fire = 100%.

package/resources/skills/skills/eu-funding/ted-tender-search/scripts/ted-search.sh

@@ -0,0 +1,46 @@
+#!/usr/bin/env bash
+# ted-search.sh — one-command EU tender search via the official free TED API.
+#
+# Usage:   ted-search.sh <BUYER_COUNTRY> [CPV_CODES] [SCOPE] [LIMIT]
+#   BUYER_COUNTRY  ISO 3-letter buyer country, e.g. HUN, SVK  (required)
+#   CPV_CODES      space-separated, default "72000000 48000000 73000000" (IT/sw/R&D)
+#   SCOPE          ACTIVE (default, open to bid) or ALL (historical)
+#   LIMIT          page size, default 50
+#
+# Examples:
+#   ted-search.sh HUN
+#   ted-search.sh SVK "72000000 48000000" ACTIVE 60
+set -euo pipefail
+
+COUNTRY="${1:?usage: ted-search.sh <BUYER_COUNTRY> [CPV] [SCOPE] [LIMIT]}"
+CPV="${2:-72000000 48000000 73000000}"
+SCOPE="${3:-ACTIVE}"
+LIMIT="${4:-50}"
+
+QUERY="classification-cpv IN (${CPV}) AND organisation-country-buyer IN (${COUNTRY})"
+
+BODY=$(python3 -c 'import json,sys
+print(json.dumps({"query":sys.argv[1],
+  "fields":["publication-number","notice-title","buyer-name","classification-cpv","notice-type"],
+  "limit":int(sys.argv[2]),"scope":sys.argv[3]}))' "$QUERY" "$LIMIT" "$SCOPE")
+
+curl -sS -m 60 -X POST "https://api.ted.europa.eu/v3/notices/search" \
+  -H "Content-Type: application/json" -d "$BODY" \
+| python3 -c '
+import json,sys
+d=json.load(sys.stdin)
+def pick(f):
+    if not isinstance(f,dict): return f or ""
+    for l in ("eng","hun"):
+        if f.get(l): v=f[l]; return v[0] if isinstance(v,list) else v
+    return next((v[0] if isinstance(v,list) else v for v in f.values()), "")
+ns=sorted(d.get("notices",[]), key=lambda n:-int(n["publication-number"].split("-")[1]))
+total=d.get("totalNoticeCount",0); shown=min(15,len(ns))
+print("TOTAL match: %d   (showing newest %d)\n" % (total, shown))
+for n in ns[:15]:
+    p=n["publication-number"]
+    title=pick(n.get("notice-title",""))[:70]
+    buyer=pick(n.get("buyer-name",""))[:55]
+    print("%s  %s" % (p, title))
+    print("   %s  https://ted.europa.eu/en/notice/%s" % (buyer, p))
+'

package/resources/skills/skills/event-design/wedding-invitations/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: wedding-invitations
-description: Use when designing a wedding invitation, save-the-date, or RSVP card from a conversation — bespoke HTML rendered to a print-ready PNG, any language, any aesthetic, fully local. Pointer to upstream wyx-sg/wedding-invitation-skill.
+description: 'Use when user says "make a wedding invitation", "design a save-the-date", "create an RSVP card", "wedding menu card", or "place cards for our wedding". Designs a bespoke wedding invitation from a conversation — bespoke HTML rendered to a print-ready PNG, any language, any aesthetic, fully local. Pointer to upstream wyx-sg/wedding-invitation-skill.'
 allowed-tools: Bash(chromium:*), Bash(google-chrome:*), Bash(microsoft-edge:*), Bash(node:*), Bash(git:*)
 ---
 

package/resources/skills/skills/github/gx-agents/SKILL.md

@@ -0,0 +1,96 @@
+---
+name: gx-agents
+category: github
+triggers:
+  - who's working on what
+  - show all agents
+  - is anyone else editing
+  - agent collision
+  - which PR is each agent shipping
+  - who owns this file
+  - agent radar
+description: >-
+  Use when several agents work in parallel and you need to see who is on which branch, worktree, or PR, or to avoid editing a file another agent already owns. Triggers: "who's working on what", "show all agents", "is anyone else editing X", "agent collision", "which PR is each agent shipping", "who owns this file", "agent radar". Drives the gx mcp tools (list_agents, who_owns, my_context) or the `gx mcp list-agents` / `gx mcp who-owns` CLI to read every agent's branch, worktree, dirty files, locks, and PR. Read-only. NOT for repo-safety repair (use gitguardex).
+---
+
+# gx agents: see every agent, avoid collisions
+
+Multiple agents (Claude, Codex) editing nearby repos step on each other: two
+edit the same file, or one edits the protected primary checkout and a later
+branch switch auto-stashes the work. `gx mcp` exposes the live picture
+gitguardex already tracks. Read it BEFORE you edit, not after the conflict.
+
+## Prereq
+
+The server ships with gitguardex (`gx mcp serve`). Register it once so any
+session can call the tools:
+
+```sh
+claude mcp add gx -s user -- gx mcp serve
+```
+
+If it is not registered, use the CLI fallback below: same data, no MCP client.
+
+## Before you edit a shared file: check ownership
+
+Ask who holds the lock before touching a path. If another branch owns it,
+coordinate or pick a different file instead of overwriting.
+
+```sh
+gx mcp who-owns path/to/file.ts
+# -> "locked by <agent> (branch agent/<name>/...)"  or  "unclaimed"
+```
+
+MCP tool form: call `who_owns` with `{ "file": "path/to/file.ts" }`.
+
+## See the whole field
+
+List every active agent lane across all your repos: branch, worktree, the
+files each is editing right now (`dirty`), held locks, last commit, and PR.
+
+```sh
+gx mcp list-agents            # human view
+gx mcp list-agents --json     # machine view (pass --no-prs to skip gh/network)
+```
+
+MCP tool form: call `list_agents` (pass `{ "include_prs": true }` when you need
+PR state, which is off by default to keep the cross-repo scan fast).
+
+Watch for `⚠ ON PRIMARY CHECKOUT`: that lane is editing the protected base, not
+an isolated worktree, so its work can be auto-stashed. Tell it to move with
+`gx branch start`.
+
+## Know your own lane
+
+```sh
+# MCP tool: my_context  -> repo, branch, worktree, onPrimaryCheckout, dirty, locks, PR
+```
+
+## Example
+
+User: "before I touch the checkout component, is anyone else on it?"
+
+```sh
+gx mcp who-owns apps/storefront/src/checkout-delivery-step.tsx
+# -> locked by codex (branch agent/codex/checkout-copy-fix)
+gx mcp list-agents --no-prs | grep -A3 checkout
+# -> codex  agent/codex/checkout-copy-fix   editing 4 file(s)
+```
+
+Conclusion: codex owns it and is editing it right now. Pick different work or
+post a handoff instead of overwriting.
+
+## Rules
+
+- **Read-only.** These tools never change a repo. To actually reserve a file,
+  use `gx locks claim --branch "<branch>" <file>`; to coordinate live tasks,
+  use Colony. `gx mcp` only shows the state.
+- **Locks lag edits.** File locks are written at commit time, so `who_owns` can
+  return `unclaimed` while another lane is mid-edit. Cross-check the lane's
+  `dirty` files in `list_agents` for in-progress work.
+- **One agent per file.** If `who_owns` (or another lane's `dirty`) shows the
+  file is taken, do not overwrite. Post a handoff or pick different work.
+- **Act on the primary-checkout warning.** A lane on the primary checkout is the
+  collision waiting to happen; route it to `gx branch start` first.
+- **This is not gitguardex repair.** For broken repo safety (dirty worktree,
+  stuck finish, `gx doctor`) use the `gitguardex` skill instead.

package/resources/skills/skills/gstack/design-shotgun/SKILL.md

@@ -2,7 +2,7 @@
 name: design-shotgun
 preamble-tier: 2
 version: 1.0.0
-description: Design shotgun: generate multiple AI design variants, open a comparison board, collect structured feedback, and iterate. (gstack)
+description: Generate multiple AI design variants at once, open a side-by-side comparison board, collect structured feedback, and iterate toward a winner. Use when the user says "explore design variants", "show me design options", "visual design brainstorm", "design shotgun", or "give me a few directions" for a UI, page, or component. NOT for auditing a single existing design — use design-review for that.
 triggers:
   - explore design variants
   - show me design options

package/resources/skills/skills/marketing/ab-test-analyzer/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: ab-test-analyzer
-description: Statistical significance calculator for A/B test results with sample size requirements, segment breakdowns, and hypothesis generation. Use when feeding test results, checking statistical significance, calculating sample sizes, analyzing experiment outcomes, or generating next test ideas based on results.
+description: 'Use when user says "analyze this a/b test", "is this result significant", "check statistical significance", "what sample size do i need", "is this a winner". Statistical significance calculator for A/B test results with sample size requirements, segment breakdowns, and hypothesis generation. Feed test results to check significance, calculate sample sizes, analyze experiment outcomes, or generate next test ideas based on results.'
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/ab-test-setup-and-analysis/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: ab-test-setup-and-analysis
-description: Designs statistically valid split tests for ads, audiences, landing pages, or bid strategies. Calculates required sample sizes before you start, monitors results during the test, and calls winners when statistical significance is actually reached — not when you feel like one is winning.
+description: 'Use when user says "set up an a/b test", "design a split test", "how long should i run this test", "calculate sample size", "can i call a winner yet". Designs statistically valid split tests for ads, audiences, landing pages, or bid strategies. Calculates required sample sizes before you start, monitors results during the test, and calls winners when statistical significance is actually reached — not when you feel like one is winning.'
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/account-structure-review/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: account-structure-review
-description: Evaluates your campaign and ad set structure against your actual goals and budget. Flags over-segmentation that fragments your data, under-segmentation that hides performance differences, budget allocation issues, and consolidation opportunities that would improve algorithmic delivery and your ability to optimize.
+description: Use when user says "review my account structure", "audit my campaign structure", "am I over-segmented", "consolidate my campaigns", "too many campaigns", "fix my ad account structure". Evaluates your campaign and ad set structure against your actual goals and budget. Flags over-segmentation that fragments your data, under-segmentation that hides performance differences, budget allocation issues, and consolidation opportunities that would improve algorithmic delivery and your ability to optimize.
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/ad-copy-variant-generator/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: ad-copy-variant-generator
-description: Analyzes your top performing ads, identifies what's working in the hooks, CTAs, messaging angles, and formats, then generates new variants that follow the same winning patterns while introducing enough variation to test meaningfully.
+description: Use when user says "generate ad copy variants", "write new ad variations", "give me ad copy to test", "more variants of my best ad", "RSA headlines", "fresh ad creative". Analyzes your top performing ads, identifies what's working in the hooks, CTAs, messaging angles, and formats, then generates new variants that follow the same winning patterns while introducing enough variation to test meaningfully.
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/ad-extension-audit/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: ad-extension-audit
-description: Reviews all your Google Ads extensions — sitelinks, callouts, structured snippets, call extensions, image extensions, price extensions — across every campaign. Flags what's missing, what's underperforming, what's outdated, and writes replacements based on your best performing ads and landing pages.
+description: Use when user says "audit my ad extensions", "check my sitelinks", "review my callouts", "are my extensions outdated", "improve my ad assets", "fix my Google Ads extensions". Reviews all your Google Ads extensions — sitelinks, callouts, structured snippets, call extensions, image extensions, price extensions — across every campaign. Flags what's missing, what's underperforming, what's outdated, and writes replacements based on your best performing ads and landing pages.
 metadata:
   platform: Google
 ---

package/resources/skills/skills/marketing/ad-spend-allocator/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: ad-spend-allocator
-description: Multi-channel budget optimization using MER, marginal ROAS, and diminishing returns analysis. Use when pasting multi-channel spend and results data, requesting reallocation recommendations, analyzing budget shift priorities, or optimizing marketing efficiency across Google, Meta, TikTok, and other channels.
+description: Use when user says "allocate my ad budget", "how should I split my spend", "reallocate budget across channels", "optimize my marketing budget", "where should I move spend", "budget shift recommendations". Multi-channel budget optimization using MER, marginal ROAS, and diminishing returns analysis across Google, Meta, TikTok, and other channels.
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/anomaly-detection/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: anomaly-detection
-description: Catches unusual performance changes across your accounts — CPC spikes, CVR drops, spend surges, impression collapses, CTR shifts — and flags them with context about what likely changed. The goal is to catch problems in hours instead of discovering them days later during a routine check.
+description: Use when user says "why did my CPC spike", "spot anomalies in my account", "what changed in my campaigns", "my conversions dropped", "detect performance issues", "flag unusual spend". Catches unusual performance changes across your accounts — CPC spikes, CVR drops, spend surges, impression collapses, CTR shifts — and flags them with context about what likely changed. The goal is to catch problems in hours instead of discovering them days later during a routine check.
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/attribution-model-comparison/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: attribution-model-comparison
-description: Runs your conversion data through different attribution models side by side — last click, first click, linear, time decay, position based, and data-driven. Shows you how credit shifts between campaigns depending on the model so you can make better budget decisions instead of over-investing in last-touch campaigns.
+description: Use when user says "compare attribution models", "last click vs first click", "which attribution model should I use", "am I over-crediting brand", "check my attribution", "compare conversion credit". Runs your conversion data through different attribution models side by side — last click, first click, linear, time decay, position based, and data-driven. Shows you how credit shifts between campaigns depending on the model so you can make better budget decisions instead of over-investing in last-touch campaigns.
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/audience-overlap-analysis/SKILL.md

@@ -1,6 +1,12 @@
 ---
 name: audience-overlap-analysis
-description: Compares your Meta ad sets and identifies where audiences overlap significantly, causing your ads to compete against each other in the same auctions. Tells you exactly which ad sets are cannibalizing each other and how much it's costing you in inflated CPMs.
+description: >-
+  Use when user says "check audience overlap", "are my ad sets competing",
+  "why are my CPMs high", "find audience cannibalization", "meta ad set overlap".
+  Compares your Meta ad sets and identifies where audiences overlap significantly,
+  causing your ads to compete against each other in the same auctions. Tells you
+  exactly which ad sets are cannibalizing each other and how much it's costing you
+  in inflated CPMs.
 metadata:
   platform: Meta
 ---

package/resources/skills/skills/marketing/bid-strategy-recommendations/SKILL.md

@@ -1,6 +1,12 @@
 ---
 name: bid-strategy-recommendations
-description: Analyzes your campaign history, conversion volume, CPA targets, and auction dynamics, then recommends the right bid strategy for each campaign — manual CPC, target CPA, target ROAS, maximize conversions, or maximize conversion value. Not a blanket recommendation, but campaign-by-campaign based on the data.
+description: >-
+  Use when user says "which bid strategy should I use", "recommend a bid strategy",
+  "target cpa or maximize conversions", "should I switch to tROAS", "review my bidding".
+  Analyzes your campaign history, conversion volume, CPA targets, and auction dynamics,
+  then recommends the right bid strategy for each campaign — manual CPC, target CPA,
+  target ROAS, maximize conversions, or maximize conversion value. Not a blanket
+  recommendation, but campaign-by-campaign based on the data.
 metadata:
   platform: Google
 ---

package/resources/skills/skills/marketing/budget-scenario-planner/SKILL.md

@@ -1,6 +1,11 @@
 ---
 name: budget-scenario-planner
-description: Models what happens to your CPA, ROAS, conversion volume, and impression share when you increase or decrease budget by any amount. Uses your actual account data and historical diminishing returns patterns, not generic industry assumptions.
+description: >-
+  Use when user says "what if I increase my budget", "model a budget change",
+  "should I scale spend", "project cpa at higher budget", "budget scenario".
+  Models what happens to your CPA, ROAS, conversion volume, and impression share
+  when you increase or decrease budget by any amount. Uses your actual account data
+  and historical diminishing returns patterns, not generic industry assumptions.
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/campaign-naming-convention-builder/SKILL.md

@@ -1,6 +1,12 @@
 ---
 name: campaign-naming-convention-builder
-description: Builds a consistent, filterable naming convention across your Google and Meta accounts based on your campaign types, objectives, targeting, and reporting needs. Makes filtering, reporting, and cross-platform analysis actually work instead of guessing what "Campaign_v3_Final_NEW" means.
+description: >-
+  Use when user says "build a campaign naming convention", "clean up my campaign names",
+  "standardize naming", "naming taxonomy for ads", "fix messy campaign names".
+  Builds a consistent, filterable naming convention across your Google and Meta
+  accounts based on your campaign types, objectives, targeting, and reporting needs.
+  Makes filtering, reporting, and cross-platform analysis actually work instead of
+  guessing what "Campaign_v3_Final_NEW" means.
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/channel-mix-optimizer/SKILL.md

@@ -1,6 +1,12 @@
 ---
 name: channel-mix-optimizer
-description: Given your total budget, recommends the optimal split across Google Search, PMax, Meta prospecting, Meta retargeting, and any other active channels based on your last 60-90 days of marginal ROAS and CPA by channel. Tells you where each additional dollar produces the best return.
+description: >-
+  Use when user says "optimize my channel mix", "how should I split my budget",
+  "best budget allocation across channels", "where should I spend more", "reallocate ad spend".
+  Given your total budget, recommends the optimal split across Google Search, PMax,
+  Meta prospecting, Meta retargeting, and any other active channels based on your last
+  60-90 days of marginal ROAS and CPA by channel. Tells you where each additional
+  dollar produces the best return.
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/client-report-narratives/SKILL.md

@@ -1,6 +1,11 @@
 ---
 name: client-report-narratives
-description: Takes your raw campaign performance data and writes the executive summary paragraph that goes at the top of the report. The plain English explanation of what happened, why it happened, and what's being done about it. The part every client actually reads.
+description: >-
+  Use when user says "write the client report summary", "write an executive summary",
+  "explain these results to the client", "turn this data into a narrative", "report intro paragraph".
+  Takes your raw campaign performance data and writes the executive summary paragraph
+  that goes at the top of the report. The plain English explanation of what happened,
+  why it happened, and what's being done about it. The part every client actually reads.
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/competitor-creative-analysis/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: competitor-creative-analysis
-description: Pulls competitor ads from Meta Ad Library and Google Ads Transparency Center, categorizes their messaging angles, formats, CTAs, and creative types, then identifies gaps in their approach you can exploit and patterns worth testing in your own campaigns.
+description: 'Use when user says "analyze competitor ads", "competitor creative analysis", "what ads are competitors running", "meta ad library teardown", "find messaging gaps". Pulls competitor ads from Meta Ad Library and Google Ads Transparency Center, categorizes their messaging angles, formats, CTAs, and creative types, then identifies gaps in their approach you can exploit and patterns worth testing in your own campaigns.'
 metadata:
   platform: Meta
 ---

package/resources/skills/skills/marketing/competitor-teardown/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: competitor-teardown
-description: Systematic competitive analysis covering positioning, messaging hierarchy, objection handling, and CTA strategy from landing page URLs or screenshots. Use when pasting a competitor URL, uploading competitor screenshots, requesting positioning analysis, or needing to understand competitive messaging and differentiation strategies.
+description: 'Use when user says "teardown this competitor", "analyze their landing page", "competitor positioning analysis", "how is this competitor messaging", "score this competitor page". Systematic competitive analysis covering positioning, messaging hierarchy, objection handling, and CTA strategy from landing page URLs or screenshots. Use when pasting a competitor URL, uploading competitor screenshots, requesting positioning analysis, or needing to understand competitive messaging and differentiation strategies.'
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/content-repurposer/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: content-repurposer
-description: Transform one long-form piece into multiple platform-specific content derivatives including LinkedIn posts, tweet threads, email snippets, ad hooks, and video scripts while maintaining voice consistency. Use when given a blog post, article, or pillar content to atomize across channels.
+description: 'Use when user says "repurpose this blog post", "turn this into a twitter thread", "atomize this content", "make linkedin posts from this article", "repurpose for social". Transform one long-form piece into multiple platform-specific content derivatives including LinkedIn posts, tweet threads, email snippets, ad hooks, and video scripts while maintaining voice consistency. Use when given a blog post, article, or pillar content to atomize across channels.'
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/conversion-path-analysis/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: conversion-path-analysis
-description: Maps out how users move through your funnel from first ad click to conversion. Identifies where the biggest drop-offs happen, which campaigns contribute most at each stage, and how long the typical conversion path takes across different audience segments.
+description: 'Use when user says "analyze my conversion path", "where are users dropping off", "map my funnel", "multi-touch attribution analysis", "what is my conversion journey". Maps out how users move through your funnel from first ad click to conversion. Identifies where the biggest drop-offs happen, which campaigns contribute most at each stage, and how long the typical conversion path takes across different audience segments.'
 metadata:
   platform: Google and Meta
 ---

package/resources/skills/skills/marketing/cpa-diagnostics/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: cpa-diagnostics
-description: When your CPA spikes, Claude breaks down exactly what caused it. It looks across your campaign data and isolates the contributing factors — audience fatigue, bid landscape shifts, creative decay, landing page conversion drops, budget distribution changes, or new competitor activity.
+description: 'Use when user says "why did my CPA spike", "diagnose my CPA", "my cost per acquisition jumped", "why did performance drop", "what is driving up my CPA". When your CPA spikes, Claude breaks down exactly what caused it. It looks across your campaign data and isolates the contributing factors — audience fatigue, bid landscape shifts, creative decay, landing page conversion drops, budget distribution changes, or new competitor activity.'
 metadata:
   platform: Google and Meta
 ---