All features in one page

Everything gadspilot does

Three distinct product layers, all sharing the same OAuth + memory + EU infrastructure. Pick what you need — they all work standalone.

MCP for Claude / ChatGPT In-app Insights Standalone AI tools

Layer 1

MCP for AI clients

3 endpoints, 106 tools, persistent memory shared across them. Compatible with Claude.ai, Claude Desktop, Claude Code, ChatGPT, Cursor and any MCP client. OAuth 2.1 — paste a URL, that's it.

Google Ads

/mcp · 76 tools

• Campaigns, ad groups, keywords, search terms
• Geographic, device, hourly breakdowns
• Auction insights, change history, recommendations
• 43 write tools (create / rename Search + Display ads, ad groups, RSAs, budget, bidding incl. MAXIMIZE_CLICKS, network toggling, language replace, geo bid modifiers, negatives, audiences targeting + exclusion) — dry-run by default
• Granular ad group + keyword control : pause/enable ad group whole, set CPC bid per keyword, batch pause keywords without losing perf history
• Full PMax management : shared sets, asset group signals, asset link/unlink (no UI needed)
• Display Remarketing : create_display_ad with auto image upload, audience exclusion, network opt-out

See all 76 tools →

Search Console

/mcp-gsc · 16 tools

• Search analytics (queries, pages, devices, countries)
• Top queries / pages shortcuts
• Sitemap list / submit / delete (write)
• URL Inspection (indexing, canonical, rich results)
• Bulk Safe Browsing + PageSpeed on top pages

See all 16 tools →

Meta Ads

/mcp-meta · 14 tools

• Facebook + Instagram campaigns + ad sets
• Insights at campaign / adset / ad level
• Breakdown by age, gender, placement, device
• 4 write tools (status, budget cents, pause)
• Multi ad-account in 1 endpoint

See all 14 tools →

🧠 Persistent memory

3 tools shared across endpoints — get_notes, add_note, get_local_change_log. Claude remembers what changed and why.

✍ Dry-run safety

Every write tool defaults to dry_run=true. Real execution requires explicit confirmation. All attempts logged.

🎯 Operating rules

8 rules injected per endpoint: context-first, write discipline, numbers discipline. AI behaves like a senior operator.

Layer 1.5 — Killer feature

Performance Max — fully scriptable via Claude

PMax is Google's hardest campaign type to manage : opaque algo, no keyword targeting, shared sets only for negatives, asset_group_signal as the only "targeting". Most MCP tools out there don't even READ this stuff. We read AND write the full PMax surface — 16 dedicated tools, no UI Google Ads needed.

🎯

Asset Group Signals (the PMax targeting layer)

• asset_group_signal_add — add search themes + audience signals
• asset_group_signal_remove — remove via resource_name
• list_audiences + get_audience — discover what's available
• audience_create_custom_segment — build "Concurrents X" from URLs/keywords

🚫

PMax Negatives (shared sets, the only way)

• list_shared_sets + list_shared_set_criteria — read existing
• shared_set_create — create NEGATIVE_KEYWORDS / BRANDS / NEGATIVE_PLACEMENTS
• campaign_shared_set_link/unlink — attach to a campaign
• shared_set_criterion_add/remove — manage list contents
• add_negative_placements — block sites/apps/YouTube channels

🎨

Asset & creative management

• asset_create — create text/image/YouTube video asset
• asset_group_asset_add/remove — link/unlink to a group with field_type
• View asset performance ratings (BEST/GOOD/LOW) in Insights tab

⚙️

Operational tools (cross campaign types)

• add_positive_locations — target FR/BE instead of excluding 200 countries
• ad_schedule_set — B2B Mon-Fri 9-19h scheduling
• conversion_action_create/update — manage goals (Lead/Purchase/Signup)
• bulk_remove_criteria — wipe N criteria in 1 batched call

What this means in practice : for the first time, you can run a complete PMax workflow inside Claude in 1 minute : "audit this PMax campaign, find search terms wasting budget, add them as negatives in the right shared set, rebalance the asset group signals based on the lowest-performing assets". Adspirer / Markifact / official Google MCP : none of them can do this — they're either read-only or don't cover the PMax mechanisms.

Layer 2

In-app Insights dashboards

Dedicated UI pages — no MCP needed. Real-time charts, KPIs, AI chat with context. For when you want a visual answer right now instead of opening Claude.

📊

Google Ads Insights

/accounts/{id}/insights

Per-account dashboard with performance chart (current vs prior period), per-campaign cards with health indicators, change history grouped by event type, dedicated tabs for Diffusion (PMax / Demand Gen / Search), audiences, demographics. AI chat sidebar with full account context.

Read the docs →

🎯

Sales lever

Account Targets

/accounts/{id}/targets

Set CPA / ROAS / monthly budget cap per account or per campaign — the AI uses them in every answer, every audit, every write proposal. Stop getting "your ROAS is 1.8" → start getting "your ROAS is 1.8 vs your 2.5 target = 28% under, top 3 contributing campaigns are X / Y / Z".

Read the docs →

🔍

Search Console Insights

/gsc/{id}/insights

4 KPIs (clicks, impressions, CTR, weighted position) with delta vs prior 30d. Chart with 2 tabs (Clicks×Impressions / CTR×Position). Top 50 queries + top 50 pages tables. AI chat specialized in SEO patterns (position 8-15 quick wins, cannibalization, CTR optim).

Read the docs →

📱

Meta Ads Insights

/meta/{id}/insights

Per Facebook ad-account dashboard — campaign / adset / ad performance, breakdowns by age / gender / placement. AI chat with Meta-specific intelligence (CBO vs ABO, attribution windows, audience overlap detection).

Read the docs →

Layer 3 — flagship

Standalone AI marketing tools

The reason most people switch to gadspilot. None of these exist on Adspirer, Markifact, or the official Google MCP. Open the dashboard, click a button, get a senior consultant's analysis.

🔍

Full SEO Audit (AI)

Multi-prompt audit of your GSC property — 30+ specialized prompts running in parallel, from the open-source claude-seo project (MIT).

✓ Categories: technical / content / schema / CWV / backlinks / E-E-A-T
✓ Score 0-100 + verdict per prompt + AI synthesis
✓ Export Markdown / DOCX / PDF (branded)
✓ Email with attachments to client
✓ History of last 10 runs per property
✓ Real OpenRouter cost shown ($0.01-0.05/run)

Full docs →

📈

Algorithm Updates Tracker

Auto-synced every 24h from 3 sources (Status Dashboard + Search Central blog + SE Roundtable), deduplicated, AI-enriched. Plotted directly on your traffic chart.

✓ Types tracked: Core, Spam, HCU, Reviews, Site Reputation
✓ Clickable G marker on the GSC chart
✓ Modal with 5-line French AI summary + sources
✓ "How to read the impact" guidance per update
✓ Direct link to Google's official announcement
✓ 24-month history backfill

Full docs →

🌐

Landing Page Analyzer

AI conversion audit on any URL. Triggered from your Ads insights (search terms, keywords, ads tables). 5 dimensions scored with concrete fixes per category.

✓ Message match (does the LP match the ad intent?)
✓ Trust signals + CTA visibility + form friction
✓ Speed via PageSpeed Insights mobile
✓ Email-ready brief for designer / dev / client
✓ History — track score evolution after redesigns
✓ 20 audits/day, 10 briefs/day per user

Full docs →

Built for client work

📨 Reports & History — ship to clients in 1 click

Every analysis you generate can be emailed to a client with branding, then is automatically archived for later retrieval. Stop juggling Notion / Slack / inbox to share results.

📤

Email reports

SEO Audit, Landing Page brief, AI chat threads — all sendable by email with editable subject + intro message. Branded PDF/DOCX attachments using your logo + name (configurable in admin). Reply-To set to your address so the client replies to you, not gadspilot.

✓ Choose which formats to attach (md / docx / pdf)
✓ CC field supported (loop in client team / colleague)
✓ 20 SEO audit emails / 10 brief emails / 10 chat emails per day

🗂

History across tools

Every audit, every landing page analysis, every algo update is timestamped and searchable. Click any past run to reload the full report + re-export with current branding. Prove value over time instead of re-running from scratch.

✓ Full SEO Audit : last 10 runs per property
✓ Landing Page audits : unlimited retention
✓ Algo Updates : 24-month auto-archived

📥

3 export formats

Markdown for Notion / GitHub, DOCX for Google Docs / Word with table of contents (via pandoc), PDF with branded cover page + score card via DomPDF. Custom reference.docx supported for full design control.

✓ UTF-8 BOM for Notepad-friendly markdown
✓ TOC auto-generated in DOCX/PDF
✓ No system deps for PDF (pure PHP)

See the agency workflow docs →

Across all layers

Built for trust + speed

🔐

OAuth 2.1 + encrypted tokens

AES-256-CBC at rest. Per-account scope. Revocable anytime.

🇪🇺

EU hosting ISO 27001

Servers in France (LWS.fr). GDPR compliant with audit trail for AI consents.

📡

Real-time, no cache

Live API every query. No 24h staleness like cached competitors.

🆓

Free during beta

No card, no usage limit. Core service stays free post-beta.

Ready to ship?

Connect your first account in under a minute. All features unlocked from day one.

Create your free account See full vs-competitors comparison →