Watch
0
0
Fork
You've already forked hyperhive
0
hyperhive/hive-agent-sock/README.md
atlas b31db6c7d9 subagent, bash: run each subagent run and bash task in its own cgroup
hive-subagent-daemon and hive-bash-daemon are now delegated their cgroup
(Delegate=memory pids, DelegateSubgroup=daemon). At start each daemon
enables memory+pids below its unit and creates runs/ (subagent) or
tasks/ (bash). Each subagent run gets runs/<name>-<n>/ and each bash
task tasks/<id>/, both with memory.oom.group=1, and the daemon removes
the directory when the run ends. An OOM kill now takes one whole run
instead of one process out of the daemon's shared cgroup, and never the
daemon itself.

Subagent runs get oom_score_adj +1000; bash tasks and both daemons stay
at 0; the harness stays at -500. The unit-level OOMScoreAdjust=500 on
the subagent daemon goes: it was both the daemon's own score and a floor
no run below it could be lowered past.

Bash tasks enter their cgroup from the existing pre_exec hook, writing
to a cgroup.procs fd opened by the parent. hive-claude builds its
Command itself and exposes neither a pre_exec hook nor the child's pid,
so a subagent's claude (and an ACP agent's process) is started as the
daemon's own binary with the run's cgroup in its environment; main()
enters the cgroup, raises oom_score_adj and execs the real program, the
trampoline swarm-review-classifier already uses. Both enter before the
program executes, so nothing it forks starts outside the run's cgroup.

Without a delegated `daemon` leaf (tests, a dev shell) each daemon logs
one warning and runs everything in its own cgroup, as before.

The shared helper is hive-agent-sock's run_cgroup module, the crate
both daemons already use for what in-container plugins share.

Closes #5007
2026-10-13 12:25:27 +02:00

49 lines
2.3 KiB
Markdown

# hive-agent-sock
What's common across the in-container "agent plugins" — the matrix / bash /
subagent MCP daemons and `forge_notify` today, plus any user-configured MCP
server declared in an agent's `agent.nix`. Three things live here:
## Wire types for the in-agent socket
The socket the `hive-agent` harness serves _inside_ the container to those
local producers. Any producer can dial it and push its own todos with an
arbitrary `subsystem` marker — the wire format doesn't restrict the set.
Unlike the host-served sockets, this one **never leaves the container**.
- the loose-ends-v2 **todo** op family
- the harness-local **reminder** op family
More in-agent request families may show up over time — the socket is
deliberately named for the agent, not for the todos.
**Why in-container**: the harness owns the todo + reminder stores locally and
signals its own turn loop directly, so `hive-c0re` isn't in either path: no
broker round-trip, no long-poll, no marker files. That locality is the
point — it's also what lets these stores travel with the agent for hive
portability.
**Don't confuse this with `hive-core-agent-sock`**: that crate is the
_host_-served core↔agent protocol on `/run/hive/mcp.sock` (the harness
talking out to `hive-c0re`). This socket is purely intra-container.
## `extra_mcp`: the `services.hyperhive.agent.extraMcpServers` spec
Not a socket protocol — a shared config type. Parses
`/etc/hyperhive/extra-mcp.json` (the nix-rendered
`services.hyperhive.agent.extraMcpServers` option) and renders each
entry into the shape claude expects in a
`--mcp-config` blob. Shared by `hive-agent` (the main session's servers) and
`hive-subagent-mcp` (the subagent-eligible subset, gated on
`availableToSubagents`) — both already depend on this crate for the socket
types above, so a new plugin-facing shared type belongs here rather than in
a crate of its own.
## `run_cgroup`: one cgroup per run
`hive-bash-mcp` puts each bash task in `<unit>/tasks/<id>/` and
`hive-subagent-mcp` each subagent run in `<unit>/runs/<name>-<n>/`, both with
`memory.oom.group=1`, so the OOM killer takes one whole run and leaves the
daemon alone. Both daemons are delegated their cgroup with `Delegate=memory
pids` and `DelegateSubgroup=daemon` (`nix/agent-modules/mcp.nix`); outside such
a unit the module logs one warning and runs stay in the daemon's cgroup.