Skip to content

docs(sdk): clarify ACP context-fill metrics #922

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

docs(sdk): clarify ACP context-fill metrics #922

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
12 changes: 12 additions & 0 deletions sdk/guides/agent-acp.mdx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -368,6 +368,18 @@ uv run python examples/02_remote_agent_server/09_acp_agent_with_remote_runtime.p
On the agent-server side, the ACP-capable REST surface lives under `/api/acp/conversations`, including `POST`, `GET`, `search`, `batch get`, and `count`.
</Note>

## Context Usage and Metrics

When the ACP server sends a `UsageUpdate`, the SDK uses its `used` value for the
current context-fill metric (`per_turn_token`) and its `size` for the context
window. This also works when the server does not include token counts in its
`PromptResponse`.

Token totals from `PromptResponse.usage` remain separate accounting values: they
can include multiple model calls within one turn. If the server does not send a
`UsageUpdate`, context fill falls back to the sum of input and output tokens from
that response. See [LLM Metrics](/sdk/guides/metrics) for token and cost tracking.

## Next Steps

- **[Creating Custom Agents](/sdk/guides/agent-custom)** — Build specialized agents with custom tool sets and system prompts
Expand Down