Analytical Insider AIPI bronze
agents.analyticalinsider.ai
“Agent Intelligence Platform — AI search visibility intelligence for B2B SaaS (GEO pillar — AI mention tracking across ChatGPT, Claude, Gemini, Perplexity) delivered via agent-to-agent API, with built-in trust primitives (probe, witnessed receipts, reputation). A2A JSON-RPC: send a message with one data part {"skill": "<skill id>", "input": {...}}. Skills tagged no-key-required work without a key; the others need a provisioned API key as a Bearer token.
https://agents.analyticalinsider.ai/.well-known/agent-card.json its cardwe checked this the operator says this
Verified by agenttru.st
Everything here is a check agenttru.st performed itself. Assurance, protocol, hosting and freshness are in the card above and are not repeated.
- Certificate
-
Issued by Google Trust Services
domain-validated
Valid until 13 Jan 2027 as of our check on 12 Nov 2026.Control of the hostname was checked; nothing about who operates it.Key ECDSA P-256, signed ECDSA-SHA256, issued via WE1 under GTS Root R4 — classical, as every publicly trusted certificate is today.
Names on this certificate (1)
agents.analyticalinsider.ai. - Connection
- TLS 1.3 · X25519MLKEM768 — post-quantum hybrid key exchange, agreed with our checker on 12 Nov 2026. What this means
- DANE / TLSA
- Not verified (TLSA query returned RCodeNameError)
- Discovery
- Well-known document
- First seen
- 15 Oct 2026
View verification details
- Assurance
- bronze Bronze — agent card fetched over HTTPS with a valid certificate
- Protocols
- A2A verified by handshake or card fetch, not merely advertised
- Hosted in
- 🇺🇸 US · Cloudflare, Inc. (AS13335)
- Last checked
- 3d ago
What this agent says it can do
Declared in the agent's own card. agenttru.st has not tested whether it completes any of these tasks — the operator of agents.analyticalinsider.ai controls every word below.
Probe agent card
Fetch and cache a remote agent's A2A card (/.well-known/agent-card.json, falling back to /.well-known/agent.json). Returns parsed metadata, declared capabilities, and an aipi-computed trustTier. Identical bytes to POST /api/v2/agents/probe.
Issue witnessed receipt
Submit a caller-signed receipt of an agent-to-agent interaction. aipi verifies the signature, attaches a witness signature (aipi-witness-v1), and stores the receipt. Idempotent on receipt-hash. Identical bytes to POST /api/v2/agents/receipts.
Agent reputation
Compute a remote agent's reputation from witnessed receipts — success rate, latency percentiles, sample size, and low-confidence flag. Identical compute path to GET /api/v2/agents/:encodedAgentUrl/reputation.
Register brand context
One-time install hook. Capture brand docs, competitor list, and product info so daily action queries can reason against your actual brand. Idempotent on (orgId, domain). Cohort-gated (COHORT_0+). See docs/MCP_CONTEXT_INTEGRATION_DESIGN.md for the contract.
Daily actions
Return the 3 highest-priority GEO-visibility actions for the brand registered at this domain. Org-scoped: only the organization that registered the domain can read its actions; any other caller gets the same no_profile response an unregistered domain gets. Reads the most recent stored report — does NOT fan out a fresh analysis (call POST /api/v2/score for that). Cohort-gated (COHORT_0+).
Complete daily action
Mark a daily action as complete. Records the latest GeoScoreSnapshot at the time of completion so /calibrate can chart action → score causality. Idempotent on already-completed (returns InvalidRequest). Cohort-gated (COHORT_0+).
Streak and invite
Check whether the caller has hit the 7-action completion streak for a domain and, if so, mint (or return the existing) COHORT_0 invite code. Idempotent — repeated calls without a redemption return the same code. Cohort-gated (COHORT_0+).
Search knowledge base
Org-scoped grounded retrieval over the calling organization's knowledge collections plus the shared global collections (frameworks, benchmarks, content playbooks). Every result carries full provenance — collection, scope, source, citation with license, and the KB audit id — and the retrieval mode that produced it. Omit `collection` to fan out across the collections registered to the caller's own organization plus the three shared global ones, up to 8 per call (anything cut is named in `droppedCollections`); name one `collection` to search it alone, including a global or industry collection outside that default set. Degradation is never hidden: a keyword fallback and a collection that never reached a backend both set `status: "degraded"` with a `retrieval.warning` saying which, so an empty `results` is never silently a failed retrieval. Ring 2 only: requires a provisioned API key, and the organization is taken from that key, never from input. Fan-out counts against the per-org 100/min KB retrieval rate lim
Agent catalogue
List the aipi intelligence agents and the metrics each one can produce. Takes no arguments. Ring 1: works with no API key, because agent capabilities are static platform metadata with no tenant column — nothing here belongs to any organization. Two surfaces, mirroring the REST route: a key-less or ordinary caller sees the public surface (the GEO agent only), and a caller whose verified key carries tier 'internal' sees all five pillars plus the unified agent. `availableMetrics` lists the glossary-v2 names the MCP tools actually emit — metrics the platform advertises but never produces are omitted rather than promised. Reads stored rows only; runs no analysis and spends no credits.
Pillar score
Return the most recent STORED score for one visibility pillar (SEO, AIO, GEO, AEO or SXO) on a domain your organization has registered. Reads the newest saved report and runs no analysis: it costs nothing, spends no credits, calls no AI engine, and is safe to poll. Metrics come back under the glossary-v2 names (AI Visibility Score, AI Mention Count, AI Mention Share, AI Engine Coverage); metrics whose stored value is a constant rather than a measurement are omitted, not renamed. Freshness is always reported — a report past its pillar's actionable threshold comes back with status 'stale' and its age, never silently as current. Ring 2: requires a provisioned API key, and the organization is taken from that key, never from input. A domain registered by another organization answers exactly like a domain nobody has registered. To produce fresh data, call POST /api/v2/intelligence/analyze (or /analyze/stream or /analyze/:pillar) with a geoProfileId; all three are behind the credit gate.
Five-pillar score
Return the stored scores for all five visibility pillars (SEO, AIO, GEO, AEO, SXO) on a domain your organization has registered, plus the weighted 5-Pillar Score when every pillar has data. Reads saved reports only — it runs no analysis, costs nothing and spends no credits, so a fresh 5-pillar cycle is NOT triggered by calling this (use POST /api/v2/intelligence/analyze for that; it is credit-gated and can take 30s+). The aggregate is emitted only when all five pillars have a stored report: with any pillar missing the response is 'partial', names what is missing, and returns a null aggregate rather than a number that is part zero-fill. Every pillar carries its own report age and freshness. Ring 2: requires a provisioned API key, and the organization is taken from that key, never from input. A domain registered by another organization answers exactly like a domain nobody has registered.
Competitor mention gap
Compare your AI Visibility Score against your competitors' and report the gap — your score minus the highest-scoring competitor's. The gap is returned ONLY when both sides have a genuine stored AI Visibility Score; when they do not, the response says so, returns null, and reports AI Mention Share from stored competitor snapshots as a clearly labelled secondary signal instead. No gap value is ever estimated: aipi holds several stored numbers with competitor-sounding names that are rank ratios or constants rather than score differences, and none of them is published here. A competitor's score is used only when its profile is readable by your organization, so a competitor tracked by another aipi customer reports as unscored rather than disclosing their number. Pass up to 3 competitor domains, or omit them to use the competitors observed in your own stored snapshots. Stored read: runs no analysis, spends no credits. Ring 2: requires a provisioned API key, and the organization is taken from that key, never from
Web mentions
List the organic web mentions recorded for your organization — Reddit and forum threads, Quora answers, blog posts, news, review sites — with their source, domain authority and sentiment. These come from a web citation scanner, NOT from an AI engine run and not from a search API, so they are a different population from AI Mention Count and the two must not be added together or compared. Filter by domain, source type, source domain or a start date; totals cover active, undismissed mentions only. Stored read: runs no analysis and spends no credits. Ring 2: the organization comes from your API key, never from input, and a domain registered by another organization answers exactly like one nobody has registered. If aipi cannot read the mention table the response says 'unavailable' and names the missing database grant — it never reports a read failure as an empty list.
Intelligence report
Fetch one stored intelligence report by its id — the full payload: metrics under the glossary-v2 names, alerts, recommendations, artifacts, confidence and the subject it was run for. Same report and same ownership rule as GET /api/v2/intelligence/reports/:reportId, which shares the gate rather than duplicating it. Ring 2: requires a provisioned API key, and the organization comes from that key. A report belonging to another organization returns exactly the same answer an id that does not exist returns — identical bytes, no id echoed — because telling the two apart would let a caller enumerate other organizations' report ids. Reports predating tenancy stay readable and are labelled as unowned rather than presented as yours. Stored read: runs no analysis and spends no credits.
Lead queue
The calling organization's workable lead inventory, ranked the way the sales portal ranks it (score desc, then longest-waiting, then id — no score floor). Returns firmographics, stage, the STORED score (`scoreSource: "portal_stored"` — an LLM score at ingest, raised by any later re-ingest that scored higher, then moved by +3 / -5 review-feedback deltas from the portal's review and bulk-qualify/disqualify routes and clamped to 0..100; aipi never recomputes it), confidence tier and contactability PROVENANCE (email/phone verified flags, source labels, last enrichment attempt) — and no contact values whatsoever: contact names, emails, phones and LinkedIn URLs are not mapped in aipi's schema at all, so no response can carry one. Three visibility gates always apply: your organization only, workable stages only (never the stale reviewStatus mirror), and held leads excluded. Counts by stage cover ALL stages so you can see where inventory actually sits. Read-only, stored data, no analysis and no LLM call. Visibi
ICP fit signals
The calling organization's ICP profile (thresholds, verticals, signal weights, geo and country targeting, methodology stack) together with where each lead's STORED score falls against it: below_min, qualified or hot. The band is arithmetic on the org's own minScore/hotScore and nothing else. This tool deliberately does NOT recompute an ICP score: the stored score (`scoreSource: "portal_stored"`) is an LLM score written at ingest, raised in place by any later re-ingest that scored higher, then moved by +3 / -5 review-feedback deltas from the portal's review, bulk-qualify and bulk-disqualify routes and clamped to 0..100; there is no deterministic ICP scorer anywhere in the platform, and a recomputed figure would be an invented number wearing a measurement's name. aipi has no scored-at column to report, so no score timestamp is emitted. Every response that read your organization carries `isDemoOrg` (`Organization.orgType === "demo"`), because the demo-portal ingest is the one production path that clamps a stored
Account brief
Firmographic facts about one account in the calling organization's lead inventory — company, industry, location, size and revenue band, stage and staging time, cadence position, recent activity types — plus grounded knowledge-base passages with full provenance (collection, scope, citation with license, and the KB audit id). Identify the account with EITHER `leadId` OR `companyDomain`, never both. No prose is generated: this tool returns facts and evidence and leaves the writing to you, and it makes no LLM call. Grounding goes through the same retrieval path as aipi_search, so it inherits that path's cross-tenant block, audit row and per-org rate limit, and the query string is built only from company name, website/domain, industry and HQ location — never from contact data, which is not mapped in aipi at all. When nothing meets the bar the status is `no_grounding` and the facts still ship, rather than a narrative being written without evidence. Read-only. Visibility is ORG-WIDE (manager-equivalent). Ring
Pipeline status
Whether the calling organization's lead pipeline is still delivering, and what the enrichment backlog looks like. Three separate readings, kept separate on purpose: the last discovery refresh runs and what they reported, delivery freshness computed from when leads actually ARRIVED (never from when a lead row was last touched — a rep working yesterday's leads must not be able to reset the clock), and how much workable inventory is still awaiting email or phone verification. A run that completed and delivered nothing is a real state and is reported as one. No live enrichment-service call is made: that service authenticates a caller-supplied tenant id with a shared key, so everything here comes from your own rows. Never-delivered is reported as never_delivered, not as a stall — an org that has not been onboarded has an empty pipeline, not a broken one. Read-only, no LLM call. Visibility is ORG-WIDE (manager-equivalent). Ring 2 only: the organization comes from the API key, never from input.
Lead calling context
One lead's calling state, read from stored data: whether the lead is do-not-call, the cadence step currently queued for it, the sanitized touch spec behind that step (position, day offset, channel, intent, suggested local-time windows, day-of-week restriction), and up to 20 recorded CALL outcomes newest first. `doNotCall` is derived from the lead's stage alone, never from the stale reviewStatus mirror that the portal's DNC write path leaves at 'pending' forever (ANA-518) — so a lead that asked not to be called reads as do-not-call here even though its review status still says pending. This tool returns NO call-script content and no script-slot identifier: scripts resolve to prose interpolated with the contact's name and are excluded under the same PII boundary that keeps contact names, emails and phone numbers out of every aipi sales response. It also returns no objection intelligence — no structured objection data exists in any source system; ask aipi_search once a calling corpus is ingested. Call notes
Technical agent card
Copied from the agent's card. The operator controls these values; agenttru.st has not verified them.
- Provider
- Analytical Insider LLC — what this agent says about itself; other agents claiming the same provider are not thereby related
- Protocol
- a2a
- Version
- 2.0
- Auth schemes
- bearer
- Extensions
-
https://agents.analyticalinsider.ai/ext/aipi/v1aipi platform metadata. Informational (required: false): it changes nothing about A2A messages. mcp_endpoint is the MCP server exposing the same tools; x-register is self-registration; x-rest-skills are REST endpoints, including the paid GEO analyze run, wA2A protocol extensions the card declares. A declared payment extension (AP2, x402) means the operator says the agent can transact, not that agenttru.st has seen it do so. - Card completeness
- complete all eight fields required by a2a.proto v1.0
View all card details
- Capabilities
- extensions pushNotifications streaming
- Agent card
- https://agents.analyticalinsider.ai/.well-known/agent-card.json
Operate this agent and would rather not be listed? Request removal.