Skip to content

docs: add Custom Tool Capability Recipe to KOTLIN_API.md #32

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
ashish-babu-03 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

docs: add Custom Tool Capability Recipe to KOTLIN_API.md #32

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
23 changes: 23 additions & 0 deletions docs/KOTLIN_API.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -64,7 +64,30 @@ val result = WorkflowScheduler(
println(result.processes)
println(result.syscalls)
```
## Custom Tool Capability Recipe

To give an agent access to your own tool, define a `ToolCapabilityGrant`
inside its `capability { }` block. This is the syscall boundary described
above — an agent can only call tools it has an explicit grant for, and only
up to the call limit that grant allows.

```kotlin
val analyst = agent("analyst") {
memory(memory)
capability(
tool = "my-custom-tool",
permission = ToolPermission.NETWORK,
scope = "mock://kaios/analysis",
limits = ToolCapabilityLimits(maxCalls = 1),
)
}
```

- `tool` names the exact tool this agent may call — access to any other tool is denied by default.
- `permission` categorizes the kind of access being granted (e.g. `NETWORK`, `FILE`, `ECHO`), not just whether access exists.
- `limits.maxCalls` caps how many times the agent may invoke the tool, so a runaway agent can't call it indefinitely.

See [examples/kotlin-runtime-api](../examples/kotlin-runtime-api/) for a full runnable example with multiple agents and grants.
## Expected Output Shape

The full example prints:
Expand Down