vinzenz/hyperhive
vinzenz/hyperhive
Watch
0
0
Fork
You've already forked hyperhive
 0
Code Activity
hyperhive/docs
History
Repository files (latest commit first)
Filename Latest commit message Latest commit date
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
..
agent-lifecycle hive: remove the read-only schedules page and the old schedule store 2026-10-13 01:45:12 +02:00
crates check-issue-refs: catch full forge issue URLs too, drop internal links from docs entirely 2026-09-09 21:15:28 +02:00
getting-started docs: say that re-creating an undeclared agent leaves it paused 2026-10-13 08:08:06 +02:00
integrations hive-github-notify: pick up a rotated PAT without a restart 2026-10-13 08:07:47 +02:00
networking docs: the review classifier's try-out service, its key and its subject 2026-10-12 13:11:21 +02:00
process agents: login shells read the hive facts too; test the readers 2026-10-12 10:52:21 +02:00
scheduler subagent, bash: run each subagent run and bash task in its own cgroup 2026-10-13 12:25:27 +02:00
swarm swarm-ui: the review try-out page shows and changes the review classifier's settings 2026-10-13 12:24:47 +02:00
tools hive-forge: pr edit --base retargets a PR 2026-10-13 08:37:13 +02:00
trust-boundary swarm-logs-auth: check agent tokens on the log store's query route 2026-10-13 08:38:05 +02:00
turn-loop agents: read the hive's name and domain at runtime, not from the build 2026-10-12 10:52:21 +02:00
web-ui hive: remove the read-only schedules page and the old schedule store 2026-10-13 01:45:12 +02:00
README.md docs: make review classifier page generic over swarms 2026-10-10 18:54:02 +02:00

README.md

hyperhive docs

Depth reference for hyperhive — the substrate, not the pitch (that's the top-level README / website). Every page here stands alone; pick the one matching your task rather than reading top to bottom. For the autogenerated NixOS options reference (every services.hyperhive.* / hyperhive.* option, host and agent), see the options site instead — this tree is prose, that one's generated straight from the module declarations.

Getting started

  • Bringing a fresh swarm online? → getting-started/setup.md (the hand steps after the first switch).
  • What does the dashboard look like, and how do I use it? → web-ui/ — the operator-facing starting point; its own sub-pages (shape, dashboard, agent, css-vars, terminal-rendering) go deeper into implementation.
  • What tools does an agent (or the operator) have available? → tools/ — hivectl (yours) plus every agent's MCP tool surface (bash, forge, lifecycle, matrix).

Agent lifecycle

Trust boundary & security

Accounts & integrations

  • How do per-agent forge accounts work? What does forge_notify poll, and how does it format wake messages? → integrations/forge.md (the swarm's Forgejo); tools/forge.md for the hive-forge CLI verbs agents actually call.
  • How does the matrix-tuwunel container work? Multiple accounts per agent? → integrations/matrix.md (the homeserver); tools/matrix.md for the MCP tool surface and services.hyperhive.agent.matrixAccounts.
  • How do I give an agent a GitHub account (gh + git push)? how's the PAT injected? → integrations/github.md (operator content up top; the gh/git-push + notification-poller mechanics are in a collapsed "Implementation" section at the bottom).
  • What's /knowledge? How does the swarm-wide knowledge repo sync, and how do I contribute a document? → integrations/knowledge.md.
  • What does hivectl do? Provisioning, gateway users, container shells? → tools/hivectl.md (the curated guide); tools/hivectl-cli.md for the exhaustive, autogenerated flag reference.

Networking & swarms

  • What nginx vhosts does the gateway serve? How does matrix discovery work? → networking/gateway.md.
  • How does DNS resolution work in agent containers? What's the bridge network for? → networking/network.md.
  • How do I connect two hives into a swarm? → swarm/ (peer hives, TLS trust).
  • Where do agent snapshots go? How does the swarm's btrfs receive endpoint authenticate a pushing hive? → networking/snapshot-store.md.
  • Who mints each credential, who reads it, and how does it rotate — and where's that shape headed? → swarm/credentials.md (current state, target state, and the progressive-enhancement rule); swarm/secrets.md for where each file lives today.
  • How does the secret store come up, who writes its grants, how's it sealed? → swarm/bao.md.
  • How does an agent get a prompt at set times, and why doesn't it fire twice after downtime? → swarm/schedules.md.
  • How do I set an agent's model or effort from the swarm, and what happens to its own pick? → swarm/inference.md.
  • How will a PR get checked against its repo's own yes/no rules before a full review? → swarm/review-classifier.md (planned, not implemented yet).

Scheduler, CI, observability

  • what's the job queue, as a general idea (not hive-c0re specifics)? → scheduler/jobq.md — operator-facing, no implementation detail.
  • How does the rebuild queue work? What are the concrete step kinds, queue sources, scheduler internals? → scheduler/coordinator.md.
  • How does the CI runner work? How does it register? → scheduler/ci.md.
  • How do I export Claude Code metrics (tokens, cost, tool calls) to Prometheus/Grafana? → scheduler/observability.md.

Crate reference

  • What does a specific Rust crate do, on its own terms? → crates/ — every workspace crate's own README.md, one level up from source; the crate itself is still the source of truth, this is just a walkable mirror.

Process & conventions