KonQuestSAFE
57-tool supervised Meta Ads MCP server for Claude Code. Open-core: full CRUD, multi-asset ads, targeting, diagnostics, safety gates. Premium adds 41 tools: optimization engine, vault intelligence, copy generation.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
Supervised Meta Ads Operating System for Claude Code. You can find the premium version here: https://farasokster.gumroad.com/l/konquest-meta-ads-mcp-premium
Open-Core Model
KonQuest Meta Ads MCP uses an open-core model:
Public (this repo) - 57 tools, MIT license:
- Full CRUD for campaigns, ad sets, and ads
- Multi-asset ad creation (video + static image, multi-dimension)
- Image upload and retrieval
- Campaign and ad set duplication
- Insights and bulk cross-account analytics
- Pixel and tracking diagnostics (5 tools)
- Catalog and DPA support (6 tools, including product set create/update)
- Full targeting toolkit (6 tools)
- Setup readiness checker (42+ checks with fix instructions)
- Validation pipeline, naming enforcement, post-write verification
- Safety tiers, rate limiting, rollback references
- 215 automated tests
Premium bundle - adds 41 tools:
- Advisory optimization engine (review queues, learning, experiments, budget governor, creative rotation)
- Vault intelligence and copy generation (brand voice, ICP targeting, concept selection)
- Greek language QA (Greeklish detection, orthography checks)
- Automation suite (diagnostics, bulk operations, account audit)
- Vault bootstrap (15 template files per client)
- Premium tests and evaluations
Premium bundle available separately. Public tools work fully without it.
What Makes This Different
- Production-tested across 37+ client campaigns with real budgets and real edge cases
- Not a wrapper - includes optimization engine, safety gates, validators, and rollback
- Vault-integrated - reads client intelligence (ICPs, brand voice, angles) to generate better ads
- Greek language QA built in (adaptable to any language validation)
- All ads created PAUSED - nothing goes live without explicit operator approval
- Supervised, not autonomous - operator confirms every write, every activation, every budget change
Tool Classification
a50fa308f0b2OBSERVED · 2026-10-08Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add konquest-meta-ads-mcp --env META_ACCESS_TOKEN=${META_ACCESS_TOKEN} -- None konquest-meta-ads-mcp==2.0.0Exposed tools (56)
38 read · 17 write · 1 destructive. Blast radius: 1 tool can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
bulk_rename_objects | write | |
check_token_status | read | |
create_ad_creative | write | |
create_ad_from_manifest | write | |
create_adset | write | |
create_campaign | write | |
create_multi_asset_ad | write | |
create_product_set | write | |
delete_campaign_structure | destructive | |
diagnose_pixel_on_site | read | |
discover_all_accounts | read | |
duplicate_adset | read | |
duplicate_campaign | read | |
estimate_audience_size | read | |
generate_names | read | |
get_account_info | read | |
get_account_pages | read | |
get_ad_accounts | read | |
get_ad_creatives | read | |
get_ad_details | read | |
get_ad_image | read | |
get_ads | read | |
get_adset_details | read | |
get_adsets | read | |
get_bulk_insights | read | |
get_campaign_details | read | |
get_campaigns | read | |
get_catalog_info | read | |
get_catalog_products | read | |
get_creative_details | read | |
get_event_stats | read | |
get_insights | read | |
get_instagram_identities | read | |
get_interest_suggestions | read | |
get_pixel_events | read | |
get_pixel_info | read | |
get_product_sets | read | |
list_custom_audiences | read | |
poll_video_processing | read | |
read_client_vault | read | |
resolve_page_identity | read | |
run_tracking_diagnostic | write | |
search_behaviors | read | |
search_demographics | read | |
search_geo_locations | read | |
search_interests | read | |
send_test_event | write | |
update_ad | write | |
update_ad_creative | write | |
update_adset | write | |
update_campaign | write | |
update_product_set | write | |
upload_ad_image | write | |
upload_video_asset | write | |
upload_video_resumable | write | |
validate_catalog_connections | read |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (5 observation(s))
- Network
- declared (5 observation(s))
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (3)
delete_campaign_structure
result = validate_greek_text("Ελληνικ\ufffdα", field_name="test")payload={"name": "Ελληνικ\ufffdα", "status": "PAUSED"},Gates applied: no_behavioural_pass.
a50fa308f0b2full audit observations/trust-audit/mcp-server/brandu-mos__konquest.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-08 | a50fa308f0b2 | SAFE | B | 89 | first audit |
Questions
What is the KonQuest MCP server?
57-tool supervised Meta Ads MCP server for Claude Code. Open-core: full CRUD, multi-asset ads, targeting, diagnostics, safety gates. Premium adds 41 tools: optimization engine, vault intelligence, copy generation.
What tools does KonQuest expose?
56 in total: 38 read-only, 17 that write, and 1 that can delete or overwrite (delete_campaign_structure). Every one is listed on this page with its risk.
Is KonQuest safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B. Separately from the audit: 1 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does KonQuest need?
It reads META_ACCESS_TOKEN and META_APP_SECRET from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does KonQuest run?
It speaks stdio, so it runs as a local process your client starts. It is published on PyPI as konquest-meta-ads-mcp.
How current is this page?
The grade is for one exact copy of the source (a50fa308f0b2), read on 2026-10-08. The repository is watched and re-audited when it changes.