Skip to content

docs(sdk): document launch-time skill overlays #770

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
kyletser wants to merge 1 commit into
base: main
Choose a base branch
from
Open

docs(sdk): document launch-time skill overlays #770

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
41 changes: 41 additions & 0 deletions sdk/arch/agent-server.mdx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -123,6 +123,47 @@ conversation.close()

If the server was started without `OH_SESSION_API_KEYS_0`, remove the `api_key=...` argument.

## Add Deployment Context at Launch

Services that start conversations through `POST /api/conversations` can add
deployment-owned context without changing the stored Agent Profile. Set
`agent_launch_additions` on the start request. The server applies these additions
after resolving the requested agent.

Use `skills_append` to attach complete `Skill` objects to the launched agent:

```json
{
"agent_profile_id": "00000000-0000-0000-0000-000000000000",
"workspace": {
"working_dir": "workspace/project"
},
"agent_launch_additions": {
"skills_append": [
{
"name": "deployment-automation",
"description": "Instructions for managing deployment automations.",
"content": "# Deployment Automations\nFollow the deployment procedure.",
"trigger": {
"type": "keyword",
"keywords": ["automation", "schedule"]
},
"disable_model_invocation": true
}
]
}
}
```

Existing agent skills remain authoritative when a name conflicts. Skills listed
in the resolved agent context's `disabled_skills` are not added. The additions
become part of the conversation's materialized agent and remain available after
an Agent Server restart, but the stored Agent Profile is unchanged.

Set `disable_model_invocation` to `true` for trigger-only skills that should not
be advertised for direct model invocation. Their content is still injected when
their trigger matches a user message.

## Expose It Safely

If another service runs on the same machine, keep the server bound to `127.0.0.1` and let that service connect locally.
Expand Down