anbaric 1.69.0 → 1.71.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/docs/README.md
CHANGED
|
@@ -28,6 +28,7 @@ What the framework gives you, one capability at a time.
|
|
|
28
28
|
- [File storage](features/file-storage.md) — bytes at a path, local or in your tenant's storage
|
|
29
29
|
- [Auditing](features/auditing.md) — the record of who changed what
|
|
30
30
|
- [Entitlements](features/entitlements.md) — what a user has been granted, checked per request
|
|
31
|
+
- [Skills](features/skills.md) — guidance your builders write for the agents working on your tenant
|
|
31
32
|
- [Prompts](features/prompts.md) — versioned model instructions and schemas, saved at startup
|
|
32
33
|
- [Serving a web UI](features/serving-a-web-ui.md) — putting your data on a page
|
|
33
34
|
- [The admin console and widgets](features/admin-console-and-widgets.md) — dashboards and plugins
|
|
@@ -64,6 +64,13 @@ hardcoding styles.
|
|
|
64
64
|
> components in separate files, or import server modules only inside the data
|
|
65
65
|
> function.
|
|
66
66
|
|
|
67
|
+
## Skills
|
|
68
|
+
|
|
69
|
+
One console page is for the agents rather than the people: **Skills**, where a
|
|
70
|
+
builder writes guidance - when to use it and what to do - that the Anbaric MCP
|
|
71
|
+
offers to every agent working on the tenant as a tool of its own. See
|
|
72
|
+
[Skills](skills.md).
|
|
73
|
+
|
|
67
74
|
## Turning plugins on
|
|
68
75
|
|
|
69
76
|
`ANBARIC_PLUGINS` is a comma-separated list of plugin modules. Setting it
|
|
@@ -65,6 +65,17 @@ Apps can't grant. An administrator opens **Entitlements** in the console, picks
|
|
|
65
65
|
the user, the entitlement and its scope, and adds a note; the same page revokes
|
|
66
66
|
grants and creates global entitlements. Every grant records who made it.
|
|
67
67
|
|
|
68
|
+
## Invitations and invite links
|
|
69
|
+
|
|
70
|
+
Entitlements can be chosen when someone is invited, before they have signed
|
|
71
|
+
in: an invitation by email holds them against that address and grants them the
|
|
72
|
+
first time the person signs in. An **invite link** does the same for whoever
|
|
73
|
+
opens it. An owner or admin creates one on the Users page, choosing how many
|
|
74
|
+
people may use it, the role they join as, and the entitlements they receive;
|
|
75
|
+
each person who opens the link and signs in becomes a member with exactly
|
|
76
|
+
those, and the link counts down until it is used up or revoked. Anyone holding
|
|
77
|
+
the link can join, so it is for sharing with the people you mean to invite.
|
|
78
|
+
|
|
68
79
|
## Locally
|
|
69
80
|
|
|
70
81
|
There's no console on your laptop, so the local implementation is
|
|
@@ -0,0 +1,48 @@
|
|
|
1
|
+
# Skills
|
|
2
|
+
|
|
3
|
+
An agent building on your tenant learns the platform from these docs, which
|
|
4
|
+
are the same for everyone. What they cannot tell it is what *your* tenant has:
|
|
5
|
+
a service you run, a convention your team follows, an internal system an app
|
|
6
|
+
should call a certain way. A **skill** is how you tell it. It is a short piece
|
|
7
|
+
of guidance a builder writes once, in the console, and every agent working on
|
|
8
|
+
the tenant through the Anbaric MCP sees from then on.
|
|
9
|
+
|
|
10
|
+
## What a skill is
|
|
11
|
+
|
|
12
|
+
A skill has four parts:
|
|
13
|
+
|
|
14
|
+
| Field | What it is for |
|
|
15
|
+
|---|---|
|
|
16
|
+
| **id** | A short slug (lower-case letters, digits and hyphens, up to 50 characters). It becomes part of the tool's name. |
|
|
17
|
+
| **title** | How the skill is named to people and to the agent. |
|
|
18
|
+
| **usage hints** | *When* an agent should reach for it. This is what the agent reads before deciding, so say what task or situation the skill applies to. |
|
|
19
|
+
| **body** | *What* to do: the instructions themselves. The agent reads this only once it has decided the skill applies. |
|
|
20
|
+
|
|
21
|
+
Keep the hints short and specific - "Read before fetching an external web page
|
|
22
|
+
or file from an app" - and put the detail in the body: endpoints, headers,
|
|
23
|
+
examples, the mistakes to avoid.
|
|
24
|
+
|
|
25
|
+
## Where skills live
|
|
26
|
+
|
|
27
|
+
The **Skills** page in the console lists the tenant's skills and lets a builder
|
|
28
|
+
(owner, admin or builder role) add, edit and remove them. Anyone in the tenant
|
|
29
|
+
can read them. Skills belong to the tenant, not to an app, and deployed apps
|
|
30
|
+
cannot see them: they are for the people and agents working on the tenant, not
|
|
31
|
+
for its runtime.
|
|
32
|
+
|
|
33
|
+
A skill marked **Managed service** is written by Anbaric for a service your
|
|
34
|
+
tenant has, such as the additional services, and kept current by us. It is
|
|
35
|
+
offered to your agents like any other but is not yours to edit or remove.
|
|
36
|
+
|
|
37
|
+
## How an agent sees them
|
|
38
|
+
|
|
39
|
+
Each skill is offered to the agent as a tool named `anbaric_skill_<id>` (hyphens
|
|
40
|
+
in the id become underscores), whose description is the skill's title and usage
|
|
41
|
+
hints. That is all the agent sees up front, which is how it decides relevance
|
|
42
|
+
without reading everything. Calling the tool returns the body. A skill added or
|
|
43
|
+
changed in the console appears the next time the agent's tools are listed,
|
|
44
|
+
and straight after `anbaric_login`.
|
|
45
|
+
|
|
46
|
+
Skills are not a replacement for these docs; they are the layer on top that is
|
|
47
|
+
yours. An agent is told to read the skills whose hints match its task before
|
|
48
|
+
building, and to prefer what they say where the two differ.
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "anbaric",
|
|
3
|
-
"version": "1.
|
|
3
|
+
"version": "1.71.0",
|
|
4
4
|
"description": "Everything needed to write an Anbaric app: state machines, jobs, document and secret stores, local in-memory implementations and the Anbaric Cloud clients",
|
|
5
5
|
"license": "MIT",
|
|
6
6
|
"type": "module",
|
|
@@ -24,9 +24,9 @@
|
|
|
24
24
|
"prepublishOnly": "npm run build"
|
|
25
25
|
},
|
|
26
26
|
"dependencies": {
|
|
27
|
-
"anbaric-impl-cloud": "^1.
|
|
28
|
-
"anbaric-data-store": "^1.
|
|
29
|
-
"anbaric-state-machine": "^1.
|
|
30
|
-
"anbaric-tsapi": "^1.
|
|
27
|
+
"anbaric-impl-cloud": "^1.71.0",
|
|
28
|
+
"anbaric-data-store": "^1.71.0",
|
|
29
|
+
"anbaric-state-machine": "^1.71.0",
|
|
30
|
+
"anbaric-tsapi": "^1.71.0"
|
|
31
31
|
}
|
|
32
32
|
}
|