Ads Uploader
NPM · @ADSUPLOADER/MCP · 2 COMPONENTS
Meta Ads MCP: analyze, edit and bulk launch Meta ads, with an Agent Inbox for approvals.
Available components
Last scanned (17 hours ago)
Scores higher than 56% of npm servers
Category Marketing and SEO
How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →
Supply Chain Security98
- No malware found by supply-chain analysis.Pass
- No known CVEs affecting this package version or its production dependencies.Pass
- No install/post-install scripts declared.Pass
- 30 of 92 dependencies flagged as unhealthy. View diagnostics → Partial
Provenance & Transparency35
- Source repository is publicly reachable at the declared URL. View diagnostics → Pass
- Provenance check failed: no build-provenance attestation is published. See how to fix → View diagnostics → Fail
- License check failed: the license (UNLICENSED) isn't a recognized OSI-approved license. See how to fix → Fail
- Actively maintained (last published 4 days ago).Pass
- Publishes a security disclosure policy (SECURITY.md).Pass
Schema Quality & AI Usability54
- AI-judged instruction clarity (good).Pass
- Context-footprint check failed: tool/resource definitions use about 15485 tokens (~262/item across 59 items; 59 tools + 0 resources), over budget; trim descriptions and params. See how to fix → Fail
- Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management100
- No destabilizing schema changes in the last 30 days.Pass
Tool Coverage89
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 67% of tool parameters carry a description.Partial
Tool Safety100
- No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
- All 1 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation.Pass
- An AI judge read all 60 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
- Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.Pass
Checks we ran (16)
- Malware scan
- Known CVEs
- Install scripts
- Dependency health
- Source repository
- Provenance
- License
- Maintenance
- Security disclosure
- Schema quality
- Stability
- Tool coverage
- Injection markers
- Destructive annotations
- Judged manipulation
- MCP protocol
Does not apply to this package (2)
- Endpoint SecurityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.
- Transport & ReachabilityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.
The score does not measure everything. What the score does not measure →
Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.
- 21 Nov 26 0
- Stability: 0.97 → pass security
- 20 Nov 26 0
- Stability: pass → 0.97 functional
- 17 Nov 26 +10
- Known CVEs: unverified → pass ▲ security
- Dependency health: unverified → 0.85 ▲ functional
- 16 Nov 26 −10
- Stability: pass → unverified ▼ security
- Known CVEs: pass → unverified ▼ security
- Tool safety: pass → unverified ▼ security
- Dependency health: 0.85 → unverified ▼ functional
- Capabilities: pass → unverified ▼ functional
- Tool coverage: 100 → unverified ▼ functional
- First check of Schema quality: unverified functional
- Package version: 0.4.2 → 0.4.3 functional
- 15 Nov 26 +15
- Malware scan: unverified → pass ▲ security
- 14 Nov 26 −14
- Malware scan: pass → unverified ▼ security
- Stability: 0.97 → pass security
- Schema quality: 368 → 256 ▲ functional
- Package version: 0.4.1 → 0.4.2 functional
- 12 Nov 26 −1
- Stability: pass → 0.93 functional
- 9 Nov 26 0
- Package version: 0.4.0 → 0.4.1 functional
Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.
Captured 21 Nov 2026 · Analysed npm/@adsuploader/mcp@0.4.3
Provenance No attestation
The registry publishes no build provenance for this version, so there is nothing to verify.
| Result | No attestation |
|---|---|
| Ecosystem | npm |
Background: How many MCP packages publish verified provenance →
Dependencies 92 packages
| Packages resolved | 92 |
|---|---|
| Stale | 30 |
| Tree resolution | Complete |
Background: SBOMs and build attestations, explained →
The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →
ads_query_insights Query Ad Performance Insights ~1,132
Query Ads Insights for one account, or 2 to 5 accountIds, one bounded page per account, as table rows by default. Friendly metrics such as leads, purchases, roas, results and cost_per_result cover business questions at campaign, ad set or ad level; ads_list_insight_fields finds other fields. Supports date presets and ranges, timeIncrement, breakdowns (asset, media and hourly breakdowns each alone) and filters. Use compare for vs last period or previous-year questions. Use include for status, budget, targeting or creative context instead of separate list calls. On rows with delivery, omitted counts and values are 0; without delivery they are null, and cost per result stays null for zero results. For audience segment mix (prospecting / engaged / existing customers) use breakdowns: [user_segment_key] alone, with timeIncrement weekly or monthly for share of spend over time. Reach, frequency and CTR are non-additive across rows. Use accountIds for cross-account views; never add totals across accounts. Zero-impression rows are excluded unless excludeZeroImpressions is false, and an empty result is a success. A large report returns status: pending; read it through ads_get_insights_job with job.id. Cached results carry meta.cache.ageSeconds; fresh skips the cache read. paging.after continues a page (paging.byAccount for several accounts). A 429 carries Retry-After. ROW_QUOTA_EXCEEDED carries the time the quota resets (meta.quota.resetsAt).
| Name | Type | Req | Description |
|---|---|---|---|
| accountId | string | – | Ad account ID used for the X-Ad-Account header, e.g. act_123. It is not sent in the Insights request body. |
| accountIds | array | – | 2 to 5 distinct accounts instead of accountId. Sequential, synchronous pages with each account currency and timezone. Never add totals across accounts. |
| actionBreakdowns | array | – | Defaults to action_type. Other dimensions produce actionRows/actionColumns and cannot be used with friendly metrics. action_type is always included. |
| actionTypes | array | – | Optional Meta action_type values used to select and flatten action-family fields such as actions or cost_per_action_type. |
| after | – | – | Single-account opaque cursor, or account-to-cursor map for accountIds. Use canonical act_ IDs. Omit an account for its first page; null marks an exhausted account. Keep the query unchanged. |
| attribution | object | – | Attribution controls echoed in response.meta. Omit this object to follow the account setting. |
| breakdowns | array | – | At most two breakdowns. Asset, media and hourly breakdowns must each be used alone. Use asset breakdowns for flexible/dynamic creative. user_segment_key = audience segment (prospecting / engaged / ex… |
| compare | object | – | Use for vs last period comparisons. Requires all_days and sync execution; maximum/data_maximum and non-default actionBreakdowns are unsupported. Previous values and changes appear beside each numeric… |
| datePreset | string | – | Named Meta date preset. Mutually exclusive with timeRange. Defaults to last_30d when neither date option is supplied. |
| excludeZeroImpressions | boolean | – | Defaults to true: drops zero-impression rows (adds impressions > 0 to Meta filtering). Set false to include them. Compare keeps a row if either period has impressions. |
| fields | array | – | Friendly business metrics or native fields. Use results/cost_per_result for optimization goals at campaign, ad set or ad level; missing native results can be derived from ad set setup and delivery. D… |
| filters | object | – | Optional entity ID filters. Use the filter matching the IDs you already know. |
| format | string | – | table returns row arrays in column order, including actionRows. objects returns keyed rows. |
| fresh | boolean | – | Skip the cache read but still cache the refreshed result. |
| include | array | – | Use for status, budget or creative context instead of separate list calls. Separate entities maps are linked to rows by meta.entityRefs, with summaries for targeting and creative assets. Defaults to… |
| level | string | – | Entity level for each result row. Defaults to account. |
| limit | integer | – | Maximum rows in this one bounded Meta page. Reduce it when a wide result approaches the MCP output budget. Defaults to 10 rows when include is set and limit is omitted (25 otherwise). Explicit limits… |
| mode | string | – | auto uses async for heavy queries. sync caps daily ranges at 90 days; async supports up to 400. |
| sort | object | – | Optional single-field server-side sort. |
| timeIncrement | string | – | Time bucketing for rows. all_days returns one aggregate bucket; daily is limited to 90 days in sync mode or 400 in auto/async mode. |
| timeRange | object | – | Explicit current date range, mutually exclusive with datePreset. Use compare for a previous period, previous year or custom range. |
| where | array | – | Metrics: gt/gte/lt/lte/eq/neq with numbers or in/not_in arrays. Names: eq/neq/contains/not_contains with strings; Meta rejects starts_with and in/not_in on names. Statuses: in/not_in arrays; campaign… |
No output schema declared.
No examples provided.
ads_report_issue Report Tool Issue ~191
Record a problem with these tools for the Ads Uploader team: results that contradict each other, a field that was silently ignored, a saved build display that differs from its preview, or a tool that does not do what its description says. Sends the summary, the expected and actual results and any IDs supplied, not names, emails, ad copy, tokens or other personal data; the server also redacts sensitive text. Returns { reported: true, reference, status: 'queued' }. Each call queues one report; limited to 10 reports per user per hour.
| Name | Type | Req | Description |
|---|---|---|---|
| accountId | string | – | – |
| actual | string | yes | – |
| adIds | array | – | – |
| buildId | string | – | – |
| expected | string | yes | – |
| jobId | string | – | – |
| requestId | string | – | – |
| severity | string | – | – |
| tool | string | yes | – |
No output schema declared.
No examples provided.
ads_request_feature Request a Missing Capability ~154
Record a capability these tools do not offer, for the Ads Uploader team: a missing tool, field, filter, action or level. Sends the goal, the missing capability, the tools tried and any IDs supplied, not names, emails, ad copy, tokens or other personal data. Similar requests are grouped and counted. ads_report_issue records problems with existing tools. Returns { requested: true, reference, status: 'recorded' }. Limited to 20 requests per user per hour.
| Name | Type | Req | Description |
|---|---|---|---|
| accountId | string | – | – |
| goal | string | yes | – |
| level | string | – | – |
| missingCapability | string | yes | – |
| toolsTried | array | – | – |
| workaround | string | – | – |
No output schema declared.
No examples provided.
ads_save_build Save Build ~346
Save a durable build: a launch configuration the web uploader can open and ads_create can launch by buildId. A spec-only build is the supported shape for agent-created configurations; the web uploader generates its own webState when it opens the build, and webState is only accepted unchanged from a build that was read. accountId can be omitted when the spec or webState carries it. Returns the build metadata with its url, launchable and launchBlockers, and display: totalAdSets, totalAds and the ads in each ad set as the web uploader renders them (display.complete false carries a reason, and display.inheritedFromSource lists fields that come from the source ad or preset). launchable describes headless ads_create launches only; any build opens in the web uploader. A saved build uses the campaign and ad set of its template ad, or new ones. Video items take the numeric Facebook video id (videoId in the upload result) as mediaId. Each call saves a new build; ads_update_build changes an existing one.
| Name | Type | Req | Description |
|---|---|---|---|
| accountId | string | – | Optional ad account ID. The server can also derive it from webState.selectedAccount.accountId, webState.selectedAccount.account_id, or spec.accountId. |
| name | string | – | Optional durable build name. |
| notes | string | – | Optional free-form notes stored on the build. |
| spec | object | – | Portable launch specification. Spec-only creation is supported and preferred for agent-created builds. |
| webState | object | – | Opaque web uploader session snapshot. Do not construct it by hand. Only provide webState when round-tripping a snapshot previously read from a build without changing its structure. |
No output schema declared.
No examples provided.
ads_save_preset Save Preset ~87
Save an existing ad as a reusable API preset. A preset name that already exists on the account is refused.
| Name | Type | Req | Description |
|---|---|---|---|
| accountId | string | yes | Ad account ID, e.g. act_123. |
| fromAdId | string | yes | Existing Meta ad ID to save as a preset. |
| name | string | yes | Preset name. |
| shareWithTeam | boolean | – | – |
No output schema declared.
No examples provided.
ads_search_targeting Search Targeting Options ~187
Search Meta targeting values for ads_create, a saved build spec, or set_targeting in ads_edit_preview. Use type=city to find targeting.cities[].key values, type=zip to find targeting.zips[].key values, or type=detailed for {id,name,type} entries. Builds use targeting.detailedTargetingGroups; edits use add/remove with the returned type and {id}. IDs must never be guessed; use this tool to discover them.
| Name | Type | Req | Description |
|---|---|---|---|
| accountId | string | yes | Ad account ID, e.g. act_123. |
| limit | integer | – | – |
| query | string | yes | City name, zip/postal code, or detailed-targeting keyword to search. |
| type | string | yes | Use city to find targeting.cities[].key values, zip to find targeting.zips[].key values, or detailed to find detailedTargetingGroups selections. |
No output schema declared.
No examples provided.
ads_update_build Update Saved Build ~554
Update one saved build by ID: rename it, edit notes, patch its portable spec with specPatch, or replace the full spec. specPatch merges field by field: omitted fields survive, nested text entries merge, supplied arrays replace and null removes a field. It needs expectedRevision, the revision from ads_get_build. Changes to spec, webState, mediaItems or uploadBatchId bump the revision. A rename or notes-only update does not bump it. Parallel updates to one build collide on the revision: a stale expectedRevision returns "Build changed since it was read" (HTTP 409), and ads_get_build returns the current revision. spec replaces the whole portable spec. webState is only accepted unchanged from a build that was read; the web uploader regenerates it when it next opens the build, and webStateStale: true until then is normal. Setting a name turns an autosaved build into a kept, named build. An open uploader tab picks up the update automatically. Returns updated metadata and display: totalAdSets, totalAds and the ads in each ad set as the web uploader renders them (display.complete false carries a reason, and display.inheritedFromSource lists fields that come from the source ad or preset); ads_get_build returns the stored state. Uploaded-media texts.perAd keys are file names or media:<uploadId>. CTA fields take Meta enum values such as SHOP_NOW.
| Name | Type | Req | Description |
|---|---|---|---|
| buildId | string | yes | Saved build ID, e.g. build_abc123... |
| expectedRevision | integer | – | Revision returned by ads_get_build. Required with specPatch and recommended for every update; the update fails instead of overwriting a newer revision. |
| name | string | – | New build name. Setting a name marks the build as manually saved, so it stops rotating out with autosaves. |
| notes | string | – | Free-form notes stored on the build. |
| spec | – | – | Replacement portable spec (the launch configuration used by ads_create with buildId). Send the FULL updated spec object, not a partial diff. Set to null to clear it. |
| specPatch | object | – | Preferred for content edits. Field-aware merge patch applied to the stored portable spec: omitted fields survive, nested objects (including individual texts.perAd/perAdset entries) merge, supplied ar… |
| webState | object | – | Opaque web uploader session snapshot. Do not construct this object by hand. Only send it when round-tripping a webState previously read from the build without changing its structure. For content chan… |
No output schema declared.
No examples provided.
ads_upload Upload Media ~309
Upload media to an ad account. Use paths for local disk media, files[].url for public URLs or Google Drive file links, or driveFolderUrl for a public Drive folder. URL and Drive modes return a background jobId to poll with ads_get_upload. Each call starts a new batch. Base64 is only for a couple of small images; presigned mode is only for byte-capable environments. In presigned mode, script:true is recommended for agents with a shell.
| Name | Type | Req | Description |
|---|---|---|---|
| accountId | string | yes | Ad account ID, e.g. act_123. |
| batchId | string | – | Optional existing batch ID for grouping with an earlier upload. |
| cwd | string | – | Optional base directory for resolving relative paths. |
| driveFolderUrl | string | – | Public Google Drive folder share link for server-side background ingestion. |
| files | array | – | Public media URLs (including Drive file links), small inline images, or presigned metadata. |
| inlineScript | boolean | – | With script:true, return uploadScript inline for clients that cannot fetch scriptUrl. |
| paths | array | – | Local files or directories to upload when running the local stdio server. Relative paths resolve from cwd. |
| presigned | boolean | – | Only for environments that can PUT bytes themselves. Large files may return multipart part URLs. |
| script | boolean | – | With presigned:true, return a short-lived scriptUrl and minimal file metadata. Recommended for agents with a shell; use inlineScript:true if downloads are unavailable. |
No output schema declared.
No examples provided.
ads_upload_finalize Finalize Media Upload ~136
Finalize a presigned upload batch after the client has PUT each file. Hosted calls return a background ingest jobId that ads_get_upload reads.
| Name | Type | Req | Description |
|---|---|---|---|
| accountId | string | yes | Ad account ID, e.g. act_123. |
| batchId | string | yes | Upload batch ID returned by ads_upload with presigned=true. |
| files | array | – | Files returned by ads_upload presigned mode after the client has PUT each file to uploadUrl or each multipart chunk to parts[].url. |
| thumbnailMap | array | – | Optional explicit video-to-thumbnail staged key map. ads_upload usually returns thumbnailFileKey on the video automatically. |
No output schema declared.
No examples provided.
ads_upload_refresh_urls Refresh Upload Links ~93
Refresh expired upload links for selected files or multipart part numbers in the same batch. Files already uploaded are kept. Single-PUT file keys may change; the returned keys are the ones ads_upload_finalize takes.
| Name | Type | Req | Description |
|---|---|---|---|
| accountId | string | yes | – |
| batchId | string | yes | Keep the original upload batch ID. |
| files | array | yes | Only files or parts whose upload links expired. Other uploads remain available. |
No output schema declared.
No examples provided.
ads_whoami Show Authenticated User ~38
Show which Ads Uploader user this session is signed in as (email, name, user ID) and the API host it talks to.
Input schema present but exposes no named parameters.
No output schema declared.
No examples provided.
How do I install the Ads Uploader MCP server?
Ads Uploader runs locally as an npm package, launched with npx -y @adsuploader/mcp. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
npm · @adsuploader/mcp
claude mcp add com-adsuploader-mcp -- npx -y @adsuploader/mcp
What is the Ads Uploader MCP server?
Ads Uploader is an MCP server listed in the public MCP registry as com.adsuploader/mcp. Meta Ads MCP: analyze, edit and bulk launch Meta ads, with an Agent Inbox for approvals. This page covers its npm package (@adsuploader/mcp).
Is the Ads Uploader MCP server safe to use?
Ads Uploader scores 77 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 21 November 2026. It declares no install or post-install scripts. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.
What tools does the Ads Uploader MCP server expose?
Ads Uploader exposes 59 tools: ads_report_issue, ads_request_feature, ads_list_catalogs, ads_list_product_sets, ads_get_product_showcase_recommendation, and 54 more. Their descriptions and schemas cost roughly 15,232 tokens of context every time the server is loaded.
Is the Ads Uploader MCP server still maintained?
Ads Uploader is still listed as active in the MCP registry. We last reached this channel on 21 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.
What licence is the Ads Uploader MCP server under?
Ads Uploader declares the UNLICENSED licence, which is not on the OSI-approved list. Read the terms before using it at work, and note this covers the source only, not the cost of any service it calls.
What did VerifyMCP test on the Ads Uploader MCP server?
VerifyMCP ran 16 checks on Ads Uploader, covering Malware scan, Known CVEs, Install scripts, Dependency health and 12 more. Endpoint Security and Transport & Reachability do not apply to an npm package, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.