curl --request GET \
--url https://api.hebbianrobotics.com/api/quality/hand-activity \
--header 'Authorization: Bearer <token>'import requests
url = "https://api.hebbianrobotics.com/api/quality/hand-activity"
headers = {"Authorization": "Bearer <token>"}
response = requests.get(url, headers=headers)
print(response.text)const options = {method: 'GET', headers: {Authorization: 'Bearer <token>'}};
fetch('https://api.hebbianrobotics.com/api/quality/hand-activity', options)
.then(res => res.json())
.then(res => console.log(res))
.catch(err => console.error(err));{
"analysis": "<string>",
"clips": 1,
"dataset": "<string>",
"measured_seconds": 123,
"segments_unassessed": 1,
"segments_unusable": 1,
"active_manipulation_share": 123,
"hands_visible_share": 123,
"two_hands_visible_share": 123
}{
"code": "<string>",
"message": "<string>"
}{
"code": "<string>",
"message": "<string>"
}{
"code": "<string>",
"message": "<string>"
}Footage-weighted hand visibility and active-manipulation shares reduced from one stored hand-activity analysis.
This endpoint performs no model call. It reads the stored span evidence and
divides its seconds by successfully answered footage, exactly as squash inspect hand-activity does. Refused and unusable windows are excluded from
the denominator and reported beside it.
curl --request GET \
--url https://api.hebbianrobotics.com/api/quality/hand-activity \
--header 'Authorization: Bearer <token>'import requests
url = "https://api.hebbianrobotics.com/api/quality/hand-activity"
headers = {"Authorization": "Bearer <token>"}
response = requests.get(url, headers=headers)
print(response.text)const options = {method: 'GET', headers: {Authorization: 'Bearer <token>'}};
fetch('https://api.hebbianrobotics.com/api/quality/hand-activity', options)
.then(res => res.json())
.then(res => console.log(res))
.catch(err => console.error(err));{
"analysis": "<string>",
"clips": 1,
"dataset": "<string>",
"measured_seconds": 123,
"segments_unassessed": 1,
"segments_unusable": 1,
"active_manipulation_share": 123,
"hands_visible_share": 123,
"two_hands_visible_share": 123
}{
"code": "<string>",
"message": "<string>"
}{
"code": "<string>",
"message": "<string>"
}{
"code": "<string>",
"message": "<string>"
}Authorizations
Capability-bearing token issued by a squash operator. Required for curation, exports, dataset registration, and job control when the deployment runs in gated mode; public reads stay open. GET /api/access reports the capabilities a given credential actually carries.
Query Parameters
Dataset or corpus id whose index root holds the analysis.
One stored hand-activity run. Omit for the most recent.
Response
Stored hand-activity shares and their denominator
Response of GET /api/quality/hand-activity: footage-weighted shares
reduced from one stored hand-activity analysis.
The shares describe only successfully answered footage. Refused and schema-invalid windows are counted separately so a caller can see when the denominator is smaller than the requested corpus.
The stored hand-activity run these figures came from — the one asked for, or the most recent when none was named. Opaque, and echoed so two answers a week apart can be told apart: same token means the corpus was not re-measured between them.
Stored clip artifacts in the analysis.
x >= 0Successfully answered footage: the denominator of every share.
Segments that could not be assessed, excluded from measured_seconds.
x >= 0Segments whose assessment was unusable, also excluded.
x >= 0Share of measured footage showing active manipulation.
Share of measured footage with at least one camera-wearer hand visible.
Share of measured footage with both camera-wearer hands visible.
