Included as a digital-experience investigation surface whose hosted MCP moves from aggregate metrics and funnels into matching sessions, chronological event transcripts, rendered replay screenshots, accessibility trees, state diffs, page definitions, and StoryAI frustration opportunities.
Concrete capability record
A field-by-field summary of what the reviewed first-party references actually support. Publisher update dates and BYO-UI review dates are shown separately below.
Segments, metrics, funnels and funnel steps; sessions and replay URLs; chronological session events; rendered session screenshots, accessibility trees, and diffs; page definitions; frustration-signal groups and StoryAI Opportunities; users and custom events in the separate Server API.
Build and compute behavioral metrics and funnels; retrieve saved segments, metrics, funnels, pages, and matching sessions; read complete session event transcripts; visually inspect replay state through screenshots and accessibility trees; discover and classify frustration signals and retrieve StoryAI Opportunities when entitled.
The current MCP tool reference says `build_segment` persists a segment and `update_segment` saves an updated segment; it also says `build_funnel` persists a funnel. The current FAQ says segments and metrics cannot be saved programmatically and that unsaved session-created objects are discarded at session end. Segment persistence therefore remains conflicted; funnel persistence is documented by the reference and is not contradicted by the FAQ.
Fullstory MCP supports browser OAuth or an API key passed as a bearer token to the hosted endpoint; Fullstory recommends a Standard key as the lowest level that currently works. An org Admin must enable both StoryAI and the MCP toggle. StoryAI Opportunities lookup tools additionally require the StoryAI Premium SKU, and the FAQ says organizations requiring fine-grained access control are excluded from Early Access.
Carry Fullstory masking, exclusion, and element-blocking rules into every investigation; scope each connection to one organization; identify whether a segment, metric, or funnel is saved or session-local; and close every `session_open` handle after visual review so server resources are released.
The MCP is beta and scoped to one Fullstory organization per connection. Group-by tables are capped at 50 rows; session tools default to 10 and hard-cap at 50. Dashboards, heatmaps, user administration, billing, privacy-rule changes, and other account settings are unsupported. Opportunity lookup needs StoryAI Premium, and beta throttling has no published numeric ceiling.
Agent access
Support: Official MCP server
Read scope: Analyze metrics, segments, funnels, pages, frustration groups, and entitled StoryAI Opportunities; retrieve matching sessions and full event transcripts; render session screenshots; inspect accessibility trees; and compare replay state over time.
Write scope: Analytics-definition persistence remains conflicted for segments: the tools reference says segment builders persist or save objects, while the FAQ says segments and metrics cannot be saved programmatically and are discarded unless a user manually saves them in Fullstory. The reference separately documents funnel persistence. No dashboard, privacy-rule, user, billing, or settings writes are supported through MCP.
Authentication: Connect to `https://api.fullstory.com/mcp/fullstory` using OAuth for per-user browser sign-in or a Fullstory API key as a bearer token for gateway or shared-service use. Fullstory recommends a Standard key for MCP and requires the organization's StoryAI and MCP feature toggles.
Approval boundary: Require administrator approval to enable StoryAI and MCP and security review before using an org-level API key for a shared gateway. Until the persistence conflict is resolved, preview any segment, metric, or funnel definition and show whether Fullstory will save it, return only a UI link, or discard it at session end.
Confirm the current tool catalog, plan, region, scopes, rate limits, terms, and write behavior before implementation.
Editorial assessment
A comparative architecture lens—not a quality score, market ranking, or buying recommendation. Scale: 1–5.
Proposed customer-shaped experiences
The output could be an export, report, graph, artifact, application, agent, workflow, or downstream feed. These proposals are derived from documented access—not claims that Fullstory ships them.
Evidence and dates
“Publisher updated” is shown only when the page exposes an update date. “BYO-UI reviewed” records when this research checked the reference. A missing publisher date is reported as missing—not replaced with the review date.
Official developer documentation confirms a beta OAuth MCP for aggregate behavioral analytics and session evidence.
Targeted first-party MCP/agent-role review; broader API inventory retained
First-party API reference or API overview recorded for this platform.
Inventory source: structurally normalized; content was not individually reopened in this pass.
First-party developer documentation or platform overview.
Inventory source: structurally normalized; content was not individually reopened in this pass.
First-party product homepage used to confirm product identity and current positioning.
Inventory source: structurally normalized; content was not individually reopened in this pass.
Tied to cited first-party evidence reviewed for this profile.
Official links recorded for deeper research but not necessarily reopened endpoint by endpoint.
System role, maturity interpretation, and architectural boundary.
Interface patterns and compositions—not vendor product claims.
