API reference · Analytics
Get cross-platform analytics
GET /v1/analytics
Read private saved reports for all collected destinations, without a 50-post sample or live provider calls. Cold or invalidated reports use a globally limited read-only builder; busy admission returns 429 with Retry-After: 60 and database failures return 503. Same-filter saved results can be retained with snapshot.stale=true. measuredAt is collection time and snapshot.generatedAt is preparation time. Production reads check the durable generation directly, without an additional edge-cache delay after collection completes. Period mode reports observed counter growth, including older content; Content mode reports lifetime performance of content published in the range. Totals, current, chart and breakdown share mode and filters. Rankings contain up to twenty destinations. POST /v1/analytics/refresh queues collection; GET that route reports progress. GET /v1/analytics/methodology defines reconciliation and freshness.
Request
Authenticate with a bearer API key. Agents calling over MCP authorise with OAuth instead and never handle the key.
curl -X GET https://api.postlake.dev/v1/analytics \
-H "Authorization: Bearer $POSTLAKE_API_KEY"Parameters
| Field | Type | Required | What it is |
|---|---|---|---|
period | string | No | Query string. |
profile | string | No | Query string. Profile username. Omit for the whole workspace. |
mode | string | No | Query string. Observed period activity, or lifetime performance of content published in the range. |
sort | string | No | Query string. Rank the complete matching destination population before limiting to 20. Unavailable readings sort last. Rates require a positive reported audience count. |
channel | string | No | Query string. Connected social account ID. Shared posts contribute only matching destinations. |
network | string | No | Query string. Optional platform name. |
from | string | No | Query string. Inclusive first UTC date, YYYY-MM-DD. |
to | string | No | Query string. Inclusive last UTC date, YYYY-MM-DD. Maximum range 365 days. |
agent | string | No | Query string. Stable creator client ID, not display name. Unknown historical attribution is not inferred. |
origin | string | No | Query string. Recorded interface, separate from creator identity. Account-wide supplements are omitted for actor/interface filters. |
Response
Success (200).
| Field | Type | Required | What it is |
|---|---|---|---|
period | string | Yes | |
since | string | Yes | |
mode | string (period, content) | No | |
until | string | No | |
measuredAt | string | No | |
snapshot | object | No | Private saved report metadata. generatedAt is report preparation, not metric collection. stale means rebuilding failed or a newer collection invalidated this report. |
snapshot.generatedAt | string | Yes | |
snapshot.stale | boolean | Yes | |
rankingScope | string | No | |
rankingMetric | string (views, likes, comments, engagement, engagementRate, impressions, clicks) | No | |
averageViews | object | No | Content mode only: lifetime views divided by destinations explicitly reporting views. Includes measured zeros, excludes unknown readings. A shared post counts once per destination. |
averageViews.value | number | Yes | |
averageViews.destinations | integer | Yes | |
nativeExtrasBasis | string | No | Supplemental native extras are lifetime counters, never period activity. |
posts | integer | Yes | Distinct posts with collected readings published in the range, not every published post. |
totals | object | Yes | |
methodologyUrl | string | No | Machine-readable metric and reconciliation definitions. |
current | object | No | Selected-mode totals for the requested window. Period: observed activity. Content: lifetime results of published content. |
previous | object | No | Same mode and filters in the preceding equal-length window. Null means no usable observations. |
timeseries | array of object | No | UTC daily observed growth in Period mode, or lifetime results by publication date in Content mode. Includes likes, comments and engagementRate. A daily rate is null without usable audience counts, not zero. |
byChannel | array of object | No | One row per actual connected social account. Account-wide insights are separate from post totals. fieldCoverage includes readings, unknown legacy readings and per-counter returned counts from latest observations scoped by observation date (Period) or publication date (Content), not historical daily coverage. |
topPosts | array of object | No | Up to 20 ranked destinations across the complete matching population, with optional thumbnailUrl. A shared post can appear more than once. |
derived | object | No | Engagement, blended engagement rate and CTR. See methodology for denominators. |
byPlatform | array of object | Yes |
Example
{
"period": "…",
"since": "2026-08-16T07:00:00Z",
"mode": "period",
"until": "…",
"measuredAt": "2026-08-16T07:00:00Z",
"snapshot": {
"generatedAt": "2026-08-16T07:00:00Z",
"stale": false
},
"rankingScope": "…",
"rankingMetric": "views"
}Limits
Reads are limited to 600 requests a minute per account. Every response carries X-RateLimit-Remaining and X-RateLimit-Reset, and a 429 includes Retry-After so a client can back off without guessing. Reads of network data (comments, followers, conversations) are metered in credits; reads of your own PostLake state are not.
Calling this from an agent
Agents do not call REST. They call tools. Over the hosted MCP server this operation is get_analytics, and the agent authorises with OAuth so your API key is never pasted into a chat or seen by the model. The reply comes back in the same normalised shape you see above, which is what lets an agent reason about it without a per-network branch.
You control what any one agent may do with it: which brands it may act for, which networks it may reach, how much it may do in a day, and whether it may spend credits you bought up front. See the MCP server.
Errors
Every failure uses one shape: a stable type, a granular code, the field at fault in param, and a plain-language fix. An agent can act on it without a person reading the message. See Errors and retries.
Idempotency-Key.Common questions
Can an AI agent do this without the REST API?
Yes. Over MCP the same job is the get_analytics tool. The agent authorises with OAuth, so no API key is pasted into a chat or seen by the model, and the answer comes back in the same normalised shape.
How often can I call this?
Reads are rate limited per account and every response carries X-RateLimit-Remaining and X-RateLimit-Reset. On a 429 the Retry-After header tells you how long to wait.
What happens when it fails?
You get one error shape: a stable type, a granular code you can branch on, the offending field in param, and a fix written so an agent can correct itself rather than stop. Full list on Errors and retries.