Watch
0
0
Fork
You've already forked hyperhive
0
hyperhive/hive-runtime/README.md
atlas cec35bfbb1 hive-runtime: compact ACP sessions through the agent's compact command
An ACP agent's session is now compacted like a claude one: proactively
once a turn crosses the percent-of-window watermark, and on the
operator's /compact or the agent's compact tool. Before, the ACP
backend's compact returned Unsupported and no watermark applied to it.

- AcpRuntime takes the same CompactionPolicy as ClaudeRuntime;
  make_session builds one PercentPolicy (with CHECKPOINT_PROMPT) and hands
  it to whichever backend runs.
- The runtime keeps the commands each session advertises in
  available_commands_update. If `compact` is among them, compaction sends
  the prompt `/compact` on the same session, which is how the ACP spec
  runs an advertised command. A proactive compaction runs the checkpoint
  turn first, as InfiniteSession does.
- With no compact command, the checkpoint turn runs, the session is
  archived, and the next turn starts a new one with the system prompt.
- Error::Unsupported had no producer left, so it and drive_turn's
  "/compact skipped" arm are gone.

Refs #4391
2026-09-30 07:41:47 +02:00

54 lines
2.6 KiB
Markdown

# hive-runtime
The layer an agent's turns are driven through: one `Runtime` interface
(`run`, `compact`, `archive`, `canceller`) with a backend per runtime.
- **claude** — `claude --print` through the `hive-claude` crate's
`InfiniteSession`. A pass-through: same spawn, same session handling, same
errors.
- **acp** — any [Agent Client Protocol](https://agentclientprotocol.com)
agent, spawned from a command, args and env handed to it (`RuntimeSpec`,
read from `HIVE_RUNTIME` / `HIVE_ACP_COMMAND` / `HIVE_ACP_ARGS` /
`HIVE_ACP_ENV`). It knows no agent by name; which agent runs, and how it is
configured, is decided in nix (`services.hyperhive.agent.runtime`,
`services.hyperhive.agent.acp.*`).
Both backends report a turn through `hive_claude::Sink` in claude's
`stream-json` shape. The ACP backend translates `session/update`
notifications into it (text and thought chunks as whole blocks, tool calls as
`tool_use` + `tool_result`, MCP tools named `mcp__<server>__<tool>`), so the
harness's stream consumers read either backend unchanged. Context usage comes
from ACP `usage_update`.
The crate depends on no hyperhive binary crate, so `hive-agent` and
`hive-subagent-mcp` can both drive turns through it.
## ACP backend: what it needs from the agent
- `mcpCapabilities.http` in its `initialize` response. The hyperhive tools
are only served over HTTP, so an agent without it is refused at startup.
- `loadSession`, to pick its session back up after a harness restart.
Without it every restart starts a new session.
## ACP backend: stopping a turn
A turn is stopped with `session/cancel`: by the `Canceller` handle, or by the
idle watchdog once no `session/update` has arrived for
`Config::idle_timeout`. An agent that has not answered the prompt 10s later
is killed, and the next turn respawns it. A watchdog stop fails the turn
with `AcpError::IdleTimeout` (`IdleKilled` if the agent was killed). This is
the only way a provider error the agent retries without reporting it, such as
an HTTP 429, ends the turn.
## ACP backend: compaction
The same `CompactionPolicy` as the claude backend decides when: after a turn
past the watermark, or on `compact` (the operator's `/compact`, the agent's
`compact` tool).
- If the agent advertises a `compact` command (`available_commands_update`),
it is sent as the prompt `/compact` on the same session, as the ACP spec
runs any advertised command. A proactive compaction runs the policy's
checkpoint turn first, as on claude.
- Otherwise the policy's checkpoint turn runs, then the session is archived,
and the next turn starts a new one, carrying the system prompt again.