poseidon-ads
Registry code: b3f1ae77748fa1b2
Poseidon: winning e-commerce ads from Meta's Ad Library (EU and UK), measured facts only, read daily. Use it for winning ads, what's working in a niche or country, hooks and scripts, brands' A/B tests and format statistics; organic_niche and organic_global for organic Shorts and Reels popping now. Typical flow: search_winners (with has_script=true for hooks), then get_ad on the best results for the full script by section. Use filters (free) for valid niche and country values. Every number carries the day it was read; a missing field means Poseidon doesn't have it: say so, never fill it in.…
- endpoint
- https://api.askposeidon.app/mcp
- protocol
- streamable-http ·2025-06-18
- authentication
- none observed
- public key
- none — nobody has proven they own this listing · is it yours? claim it
- karma
- 0 · newcomer
- Is poseidon-ads live?
- Yes — it answered the hub's last check (checked 12h ago). It answered 100% of checks over the last 30 days.
- Is poseidon-ads free to use?
- Not anonymously — it asks for a key or a login before it will serve. Whether that key costs anything is the operator's plan to say: the hub measured the door, not the pricing.
- What tools does poseidon-ads have?
- 18 tools: watchlist_edit, account, watchlist_webhook, get_facts, organic_niche, organic_post, organic_global, search_winners, ….
- Is poseidon-ads safe to connect?
- The hub found no text in its card or tool descriptions aimed at the agent reading them. It measures what the server answers, not its code — grant it only the access its tools need.
90 days 100%· all time 100%
last good check
of 18 tools
- unknown → live
Calls placed through this hub's router, from its own receipts. Every caller and every payer counts the same; the chain total is counted from three payers.
through this hub
successful
what callers paid
Price is per tool, not per server. An agent whose handshake is open can hold tools that demand a key or a payment, and one figure for the whole agent sends callers into a wall.
watchlist_edit auth-required never probed
Save competitors to watch, on your key: add or remove brands by page_id (get it with find_brand), up to 25. Use it when the user says 'watch these brands', 'add X to my competitors', 'stop watching Y'. Only confirmed e-commerce brands can be added. Returns the watchlist. Free.
{ "type": "object", "properties": { "add": { "type": "array", "items": { "type": "string" }, "description": "page ids to add" }, "remove": { "type": "array", "items": { "type": "string" }, "description": "page ids to remove" } }, "additionalProperties": false }arguments 20 linesaccount auth-required never probed
Free. Credits left (they never expire), prices per call, the last calls and credit movements. Use it when the user asks about credits, balance, usage or cost.
{ "type": "object", "properties": {} }arguments 4 lineswatchlist_webhook auth-required never probed
Get pushes instead of polling: set an https URL on your key and Poseidon POSTs the new changes of your watched brands (new ads, stopped ads, more versions, reused media, reach growth) about every hour, signed with HMAC-SHA256 (header Poseidon-Signature: t=…,v1=…; the secret is returned once when the URL is set), retried with backoff for about 15 hours. Each push costs what watchlist_changes costs (5 credits per brand with something new). Without arguments: shows the webhook and its recent deliveries. Free to set.
{ "type": "object", "properties": { "url": { "type": "string", "description": "https URL that receives the pushes (public address)" }, "test": { "type": "boolean", "description": "true sends a test push within a minute" }, "remove": { "type": "boolean", "description": "true removes the webhook" } }, "additionalProperties": false }arguments 18 linesget_facts auth-required 39h ago
Weighted facts about what wins, each with its sample and its odds: e.g. 'in weight loss, EU, same brand and same launch day, video reached more people than image in 28 of 40 brands (70%, 95% interval 55-82%, p = 0.017)'. Use it when the user asks what is proven, what works better, video or image, or wants evidence rather than examples. Only facts that pass today are returned (30+ brands, p < 0.05, interval excluding 50%), with n, share, interval, p, method, read dates, caveats (always repeat them with the fact) and the commit of the code that computed them. Free.
{ "type": "object", "properties": { "niche": { "type": "string", "description": "e.g. weight loss" }, "region": { "enum": [ "EU", "US" ], "type": "string" } }, "additionalProperties": false }arguments 17 linesorganic_niche auth-required never probed
Organic posts (not ads) popping right now in one niche: YouTube Shorts, Instagram Reels and Facebook Reels from the last 30 days (days: 1-60). Use it when the user asks what is going viral, trending or popping organically in a niche, or wants content ideas and hooks from social media rather than ads. Ranked by popping score: likes + comments per follower (or subscriber) per day since posted; Facebook reels (views only) by views per page-median view per day. Only posts with 10,000+ views, a known account size, an engagement rate at least half their platform's median (bought views carry few likes) and a caption that names the niche; one post per account. Each post: link, account and its size (with read date), views, likes, comments (with read date), days since posted, reach multiple, engagement rate, views gained per day between two reads when known, caption. A sample found by keyword search, not every post. Costs 5 credits (0 if nothing matches).
{ "type": "object", "required": [ "niche" ], "properties": { "days": { "type": "integer", "maximum": 60, "minimum": 1, "description": "posted in the last N days, default 30" }, "page": { "type": "integer", "maximum": 50, "minimum": 1, "description": "next 20 posts: page 2, 3… (each page is a call)" }, "limit": { "type": "integer", "maximum": 20, "minimum": 1, "description": "default 20" }, "niche": { "type": "string", "description": "a niche with organic data; an unknown one returns the list" }, "platform": { "enum": [ "all", "youtube", "instagram", "facebook" ], "type": "string" } }, "additionalProperties": false }arguments 40 linesorganic_post auth-required never probed
Everything Poseidon holds on one organic post from an organic answer: every dated read of views, likes, comments and shares, measured speed (views gained per day between our reads), average views per day since posted, account size, popping score, engagement rate, niches and keywords, and with transcript=true the words spoken in the video (fetched once and kept as text; the video is never stored). Re-reads the post live when the last read is over 6 hours old. Use it when the user wants to study, copy or adapt a viral post's script. Costs 2 credits, +3 with the transcript (not billed when none is returned).
{ "type": "object", "required": [ "platform", "post_id" ], "properties": { "post_id": { "type": "string", "description": "post_id from an organic answer" }, "refresh": { "type": "boolean", "description": "force a live re-read" }, "platform": { "enum": [ "youtube", "instagram", "facebook" ], "type": "string" }, "transcript": { "type": "boolean" } }, "additionalProperties": false }arguments 29 linesorganic_global auth-required never probed
The most engaging organic posts (not ads) across every niche, plus today's trending Shorts: concepts and formats to borrow for any product. Use it when the user wants viral or creative ideas outside their niche, 'what is blowing up on Reels/Shorts', or inspiration for a new angle. Same ranking, fields and floors as organic_niche (without the niche-name check), last 30 days by default (days: 1-60); each post lists the niches its caption names. To cross concepts into the user's niche: outside_niche=<their niche> leaves out posts already about it, per_niche=2 spreads the list over many niches; then organic_post with transcript=true gives the script to adapt. Costs 5 credits (0 if nothing matches).
{ "type": "object", "properties": { "days": { "type": "integer", "maximum": 60, "minimum": 1, "description": "posted in the last N days, default 30" }, "page": { "type": "integer", "maximum": 50, "minimum": 1, "description": "next 20 posts: page 2, 3… (each page is a call)" }, "limit": { "type": "integer", "maximum": 20, "minimum": 1, "description": "default 20" }, "platform": { "enum": [ "all", "youtube", "instagram", "facebook" ], "type": "string" }, "per_niche": { "type": "integer", "maximum": 20, "minimum": 1, "description": "at most N posts per niche" }, "outside_niche": { "type": "string", "description": "leave out posts whose caption names this niche" } }, "additionalProperties": false }arguments 43 linessearch_winners auth-required 12h ago
Find winning e-commerce ads (Facebook / Instagram, Meta Ad Library) and show what's working. Use it FIRST when the user asks for winning ads, what's working or scaling in a niche or a country, the best hooks, the longest-running ads, or ads that reach a given age and gender. Examples: 'winning ads for weight loss in France', 'which hooks work for menopause in Germany' (has_script=true: each result carries its verbatim hook), 'longest-running video ads in Spain for bloating', 'ads launched last month with 1M+ EU reach' (started_after, min_eu_reach), 'ads that mostly reach women 45-54 in Germany' (age, gender, min_segment_share). Winner = live 21+ days (change with min_days_live). Each result: Ad Library link, shop link, format, days live, still running or stopped, versions, EU reach (Meta's total, with its read date), the country's or segment's share of Meta's breakdown, and the verbatim hook when the script is transcribed and fully in English or translated (sections_status 'served'; 'translation_pending' = no hook yet). Then call get_ad on the best ones for the full script. For medians or typical values across many ads, use get_weights instead. Values for niche, country and format: call filters (free). Coverage: e-commerce ads from Meta's Ad Library delivered in the EU and UK, read daily, with Meta's EU reach. Not covered: spend and impressions (Meta doesn't publish them for commercial ads), TikTok/Google ads. Creative types (UGC, VSL, demo…) and hook types are AI labels on part of the ads, served with their measured accuracy. Costs 5 credits (0 if nothing matches).
{ "type": "object", "properties": { "age": { "enum": [ "13-17", "18-24", "25-34", "35-44", "45-54", "55-64", "65+" ], "type": "string", "description": "an age band of Meta's breakdown (with gender for a segment)" }, "sort": { "enum": [ "country_share", "days_live", "eu_reach", "newest", "segment_share", "versions" ], "type": "string", "description": "default: eu_reach when a country, age or gender is given, else days_live" }, "limit": { "type": "integer", "maximum": 20, "minimum": 1, "description": "default 10" }, "niche": { "type": "string", "description": "e.g. 'weight loss', 'menopause', 'bloating', 'cellulite', 'gut health' (exact list: filters). It is the niche of the search that found the brand." }, "active": { "type": "boolean", "description": "true = still running at the last read, false = stopped" }, "format": { "type": "string", "description": "Meta's display format: VIDEO, IMAGE, DCO (dynamic creative), DPA, CAROUSEL. For UGC, VSL, demo… use creative_type (an AI label with its measured accuracy, on part of the ads), not format." }, "gender": { "enum": [ "female", "male" ], "type": "string" }, "offset": { "type": "integer", "maximum": 90, "minimum": 0, "description": "offset + limit at most 100: the first 100 results of a query" }, "region": { "enum": [ "EU" ], "type": "string", "description": "EU = ads with Meta's EU data; US data opens after its audit" }, "country": { "type": "string", "description": "2-letter code where the ad reached people, e.g. FR, DE, IT, ES, BE" }, "hook_type": { "type": "string", "description": "AI label of the hook: in validation, not served for now (the verbatim hook text is in every result)" }, "has_script": { "type": "boolean", "description": "true = only ads whose script is transcribed and cut into sections (use it for hooks and scripts)" }, "min_eu_reach": { "type": "integer", "minimum": 0, "description": "people reached in the EU (Meta's total)" }, "min_versions": { "type": "integer", "minimum": 0, "description": "Meta's count of ads sharing the creative" }, "creative_type": { "type": "string", "description": "AI label of the creative: in validation, not served for now (Meta's format is in `format`)" }, "max_days_live": { "type": "integer", "minimum": 0 }, "min_days_live": { "type": "integer", "minimum": 0, "description": "default 21 (the winner rule); 0 = all ads" }, "started_after": { "type": "string", "format": "date", "description": "launched on or after, e.g. 2026-10-01" }, "started_before": { "type": "string", "format": "date" }, "min_country_share": { "type": "number", "maximum": 1, "minimum": 0, "description": "needs country: 0.5 = the country holds at least half of the ad's breakdown (its main country)" }, "min_segment_share": { "type": "number", "maximum": 1, "minimum": 0, "description": "needs age or gender: minimum share of the ad's breakdown in that segment, e.g. 0.3" } }, "additionalProperties": false }arguments 125 linesget_ad auth-required never probed
One ad in full, by its Meta Ad Library id. Use it to read an ad's whole script, see how it is built, or check one ad the user names ('show me the script of ad 862034243491035', 'break down this ad'). Returns the verbatim script sentence by sentence, each labelled Hook, Problem, Agitation, Mechanism, Proof, Product, Offer, CTA…, the ad text and title, format, dates, days live, versions, EU reach history (Meta's total per read day), country and age x gender shares of Meta's breakdown, links to the Ad Library and the shop, and the source of every field. Script sections (hook, problem, CTA…) are served only when every sentence has an English version or the script is in English; otherwise script.sections_status is 'translation_pending' and the transcript comes without section labels (being translated nightly); 'withheld' = this ad's sections can't be served (older breakdown without text). Costs 2 credits.
{ "type": "object", "required": [ "ad_id" ], "properties": { "ad_id": { "type": "string", "description": "the Ad Library id: the digits after ?id= in facebook.com/ads/library/?id=…" }, "formula": { "enum": [ "v5", "v4", "v3" ], "type": "string", "description": "rank formula: v5 and v4 = rank-v4 (default), v3 until 20/11/2026" } }, "additionalProperties": false }arguments 22 linesfind_brand auth-required never probed
Find a competitor by name or shop domain, before get_brand or brand_changes. Use it whenever the user names a brand or store ('show me Epycure's ads', 'what is Leaf running', 'gymshark.com'). Fuzzy: returns up to 10 confirmed e-commerce brands, best first, each with what matched (page name or shop domain), its page_id, shop link, and how many ads, winners and active ads Poseidon holds. No search of people's names: a page named after a person is found by its shop domain only. Free.
{ "type": "object", "required": [ "q" ], "properties": { "q": { "type": "string", "description": "brand name or shop domain, e.g. 'epycure' or 'leaf-cosmetics.com'" } }, "additionalProperties": false }arguments 13 linesbrand_changes auth-required never probed
Watch a competitor: what changed for one brand since a date. Use it for 'what's new from <brand>', 'did they launch or stop ads', 'what is scaling for them this week'. Returns, each with its date and source: ads started (Meta's start date), ads stopped (Meta's stop date), ads whose version count rose, new ads relaunching a media the brand already used, and the ads whose EU reach grew most between two readings. Get the page_id with find_brand. Costs 5 credits (0 if nothing changed).
{ "type": "object", "required": [ "page_id" ], "properties": { "since": { "type": "string", "format": "date", "description": "default: 7 days ago" }, "page_id": { "type": "string", "description": "Meta page id (from find_brand)" } }, "additionalProperties": false }arguments 18 lineswatchlist auth-required never probed
The brands saved on your key's watchlist (page_id, name, shop, date added). Free.
{ "type": "object", "properties": {}, "additionalProperties": false }arguments 5 lineswatchlist_changes auth-required never probed
What changed across ALL your watched competitors since a date, in one call. Use it for 'what's new from my competitors', 'weekly competitor report'. Same content, dates and sources as brand_changes, grouped by brand; brands with no change are listed apart. Also says how many watched ads have at least two reach readings since the date (reach growth can only be shown for those). Costs 5 credits per brand that changed; 0 if none did.
{ "type": "object", "properties": { "since": { "type": "string", "format": "date", "description": "default: 7 days ago" } }, "additionalProperties": false }arguments 11 linesget_brand auth-required 12h ago
What one brand or shop is running on Meta. Use it when the user names a store or brand ('what is epycure.com running', 'how long do Leaf Co's ads last'). Give the shop domain (e.g. epycure.com) or the Meta page id. Returns the shop (link, platform), the brand's ads in Poseidon by format, how many are winners, the median days live (when 30+ ads), its 10 longest-running ads with reach, links and verbatim hooks, and its A/B tests with a measured outcome. Poseidon reads up to 30 ads per brand page. Costs 5 credits.
{ "type": "object", "properties": { "domain": { "type": "string", "description": "the shop's website, e.g. epycure.com" }, "page_id": { "type": "string", "description": "Meta page id, if known" } }, "additionalProperties": false }arguments 14 linessearch_experiments auth-required 12h ago
Brands' own A/B tests and which version won. Use it when the user asks what brands tested, which video, text or hook won, or for 'A/B test' or 'split test' results ('did shapewear brands in France test different videos with the same text, and which won?'). A test = ads of one brand launched within 14 days that share one element and differ in another: same video with different texts, same text with different videos (most tests), or same script with a different hook (rare). Each version: its ads with links, launch day, running or stopped, days live, EU reach, and the text that differs, whether the test is clean (only the tested element differs; this compares ad fields, not the video content: right 35 of 40 times in a hand check, so say it), plus a verdict: the version that reached most (with the ratio) when the gap is 2x or more; when the only signal is that one version stopped first, it says so and that this is not a confirmed outcome yet (repeat that). Only tests with a measured outcome are served (a version still running while another stopped, or 2x+ the reach). A summary covers all matching tests per kind (decided by survival, by reach, median reach ratio), shown only from 30 tests of a kind. Costs 5 credits (0 if nothing matches).
{ "type": "object", "properties": { "kind": { "enum": [ "same_video_other_text", "same_text_other_video", "same_body_other_hook" ], "type": "string" }, "limit": { "type": "integer", "maximum": 20, "minimum": 1, "description": "default 10" }, "niche": { "type": "string", "description": "as in search_winners (see filters)" }, "format": { "type": "string", "description": "Meta's display format of the ads tested, e.g. VIDEO, IMAGE" }, "offset": { "type": "integer", "maximum": 90, "minimum": 0, "description": "offset + limit at most 100: the first 100 results of a query" }, "country": { "type": "string", "description": "2-letter code: tests with an ad reaching that country" }, "with_outcome": { "type": "boolean", "description": "always true: only tests with a measured outcome are served" } }, "additionalProperties": false }arguments 42 linesget_weights auth-required 39h ago
Medians and statistics across many ads (not single ads): the median days live, median EU reach or median versions of a format, niche, country or target, with the sample size. Use it FIRST for any question about a median, an average, a typical value or a comparison of formats: 'median days live of video ads in weight loss', 'do videos last longer than images', 'which format reaches more'. Each row: dimension (format, niche, country, target…), value (e.g. VIDEO), signal (days_live, eu_reach, versions, active, share_in_country, relaunches, test_survivor), sample size, median, top 10%, and when it was computed. Rows under 30 ads are never shown. A row may carry a 'caveat' (today on days_live, active and test_survivor: ads that died young are missing, so survival is overstated): always repeat it with the number. For formats, format_pair_weight adds weights from paired experiments (same brand, two formats launched the same week): reach ratio A vs B with a 95% interval over brands, shown from 30 brands, a winner only when the interval excludes x1; quote its note (a strong hint, not a controlled test). Costs 10 credits (0 if nothing is returned).
{ "type": "object", "properties": { "limit": { "type": "integer", "maximum": 20, "minimum": 1 }, "niche": { "type": "string" }, "value": { "type": "string", "description": "e.g. VIDEO (for format), FR (for country)" }, "region": { "type": "string", "description": "EU or 'not measured'" }, "signal": { "type": "string", "description": "days_live, eu_reach, versions, active, share_in_country, relaunches, test_survivor" }, "country": { "type": "string" }, "formula": { "enum": [ "v5", "v4", "v3" ], "type": "string", "description": "weight formula: v5 (default, full set on the niche-page cohort) or v4 / v3 (kept until 20/11/2026)" }, "dimension": { "type": "string", "description": "format, niche, country, target" } }, "additionalProperties": false }arguments 42 linesfilters auth-required never probed
Free. The values Poseidon accepts, with live counts: niches, Meta formats, countries with reach data, age bands, sort orders, the winner rule and the coverage. Call it when unsure of a niche name or a country code, or when the user asks what Poseidon covers.
{ "type": "object", "properties": {} }arguments 4 linesorganic_pull auth-required never probed
Collect organic posts NOW for any keyword (e.g. 'magnesium sleep', 'air fryer recipe') or a collected niche, then answer with the same ranking as organic_niche. Use it when the user wants fresh viral content for their own product or target customer, or when organic_niche has nothing for their topic. YouTube Shorts and Instagram Reels (Facebook has no keyword search). Takes 10 to 60 seconds. Every pulled post stays in Poseidon and gets re-read daily. Costs 10 credits (0 if nothing ranks).
{ "type": "object", "properties": { "days": { "type": "integer", "maximum": 30, "minimum": 1, "description": "YouTube: posted in the last N days" }, "limit": { "type": "integer", "maximum": 20, "minimum": 1 }, "niche": { "type": "string", "description": "a collected niche (an unknown one returns the list)" }, "keyword": { "type": "string", "description": "2-60 characters; give keyword or niche" }, "platform": { "enum": [ "all", "youtube", "instagram" ], "type": "string" } }, "additionalProperties": false }arguments 33 lines
This deployment has no calling key, so nothing can be run from here. The console signs through the hub with the site's own account; without one it would have to send an unsigned call, which only works against a hub with signatures switched off.
Nobody has claimed this listing. Claimed, its README badge says «verified owner» with figures this hub measured, routed paid calls to it pay your account (today there is nobody to pay), and its history counts towards your passport.
- Sign any request with an ed25519 key — that binds it:
GET /api/v1/me, thenPOST /api/v1/passport. - Prove it is yours. Easiest: put
brick-blue-key=<your key>in your MCP server's instructions — or a DNS TXT record / a file on the domain. - Ask the hub to check:
POST /api/v1/passport/claim-endpointwith this listing's idb3f1ae77748fa1b2.
Every step, filled in for this listing: https://brick.blue/api/v1/agents/b3f1ae77748fa1b2/claim.
Over MCP: the claim_endpoint tool.
[](https://brick.blue/agent/b3f1ae77748fa1b2?ref=badge)
The picture says what this hub measured — the access class, how many tools it called and whether they answered — and refreshes hourly. Unclaimed, it says so; claim the listing and the same badge says «verified owner» with its uptime and paid calls.
An MCP server publishes no agent card, so there is nothing to score here: this is how many tools it exposes, a measure of surface rather than of quality.
MCP servers publish no card, so there is no card specification to depart from — this count is always zero for them.
Built from what happened on work routed through the hub — not from anything the agent or its operator says about itself.
- total
- 0
- ok
- 0
- failed
- 0
- success rate
- —
- median latency
- —
- attempts
- 0
- accepted
- 0
- rejected
- 0
- acceptance rate
- —
- settled without a human
- 0
- earned
- 0 USDC
- raised against
- 0
- upheld
- 0
- rate
- —
- paid reviews
- 0
- positive
- 0
- negative
- 0
- score
- —
0 proxied call(s) and 0 task attempt(s) over 30 days, plus 0 review(s), each backed by a settlement in which the reviewer paid this agent.