优化前端,优化前端指令,优化自定义控制

This commit is contained in:
minghanchen
2026-08-24 09:40:16 +08:00
parent fff7657fea
commit 8d05815773
130 changed files with 5543 additions and 2206 deletions
@@ -3,7 +3,7 @@ kind: CapabilityCatalog
metadata:
id: driver-seat-native-can-capabilities
name: Driver seat native CAN capabilities
version: 1.0.0
version: 1.1.0
spec:
domain: seat
capabilities:
@@ -108,3 +108,40 @@ spec:
description: Bounded motion duration; omitted values default to 500 ms.
risk_level: high
confirmation_policy: on_ambiguity
- name: seat.set_driver_position_native
description: >-
Alias for recalling one of the three physical driver/front-left seat memory
slots. The complete 0x4C1/0x15 and /0x16 payloads must have been learned from
the PAD and a fresh 0x494 must prove all four axes initialized before replay.
target_devices: [driver_seat_memory]
target_scopes: [driver, front_left]
parameters:
- name: slot
type: integer
required: true
minimum: 1
maximum: 3
allowed_values: [1, 2, 3]
description: Previously learned PAD memory slot.
risk_level: high
confirmation_policy: on_ambiguity
- name: seat.recall_driver_memory_native
description: >-
Recall one of the three physical driver/front-left seat memory slots by
alternating that slot's learned 0x4C1/0x16 and /0x15 payloads for ten cycles.
0x494 does not expose final absolute motor positions, so delivery is not
represented as proof that the seat reached its final position.
target_devices: [driver_seat_memory]
target_scopes: [driver, front_left]
parameters:
- name: slot
type: integer
required: true
minimum: 1
maximum: 3
allowed_values: [1, 2, 3]
description: Previously learned PAD memory slot.
risk_level: high
confirmation_policy: on_ambiguity