a swarm o agents, each in its own nspawn cage, gossiping over unix sockets. config changes flow as git commits, the operator approves them in a browser, every deploy is a tag. cyberpunk-themed dashboard included. 💜
  • Rust 68.7%
  • Nix 15.7%
  • JavaScript 8.4%
  • CSS 3.7%
  • TypeScript 1.9%
  • Other 1.6%
Find a file
Repository files (latest commit first)
Filename Latest commit message Latest commit date
atlas d3d73b5ffb refactor(#2815): derive the transient pill from the running node
The dashboard pill was declared once per DAG at submit time, so a rebuild
reported `rebuilding` for its entire life — through the prebuild, the
stop, the swap, the tail and the reconcile. It named the intent of the
request, not what was happening.

It is now read off the nodes actually running. A node lights a pill when
it is `Running` and declares the agent's own resource. Declaring is the
test, not targeting: `Prebuild` and `MetaSync` name an agent but are
lease-exempt on purpose (the container keeps serving), so they must not
light one. It is also not the lease *owner* — `resource_state()` answers
"who holds the slot", which is a different question from "what is
running", and a descendant that borrows an ancestor's grant never
appears in that map.

`TransientKind` is gone entirely rather than being re-derived. The label
is the node's own wire tag (`NodeKind::as_str`) — the same vocabulary
`NodeView.kind` already ships, so a pill and a DAG node name an operation
identically and there is no second taxonomy to keep in step. Work with no
node behind it (destroy, migration) supplies its own literal.

`DagSpec::transient`, `Claim::transient`, `DagMeta::transient` and
`NodeKind::Dag`'s `transient` field all go with it.

## the safety half, which is deliberately not the display half

`crash_watch::is_deliberate_stop` used to match a `TransientKind` to
decide whether a vanished container was intentional or a crash. That made
a pill's display vocabulary decide an alerting question, so renaming or
adding a label would silently move the alerting boundary.

`TransientState` now carries two independent fields: `label` (rendered,
nothing branches on it) and `deliberate_stop` (read only by the crash
watcher). The producer sets the second, because the producer is the only
thing that knows — it is not recoverable from the first.

For queue work that value is `NodeKind::takes_container_down()`, and it
is emphatically not "holds a lease": `Create` and `Start` hold the
agent's lease exactly like `Stop` does, and a container dying *while
starting* is a real crash that must keep reporting as one. The default is
`false` on purpose — a wrong `false` costs a spurious crash event, a
wrong `true` swallows a real crash silently.

## known cost, accepted on the issue

A restart no longer reads `restarting`. No `NodeKind` is unique to a
restart — `restart_chain` reuses `Signal` / `StopForUpdate` / `Drain` /
`Reconcile` — because "restart" is a property of the DAG's shape, not of
any node. A restart now reads `signal` / `stop_for_update`, then the
agent returns.

`Start` / `Stop` / `PostSwap` run inside a lease-holding ancestor and
re-declare nothing, so they light no pill and the agent reads idle for
those windows. Closing that is the resources-where-constructed work
(#2818), not this change.

Checked with clippy (`--all-targets -D warnings`), `cargo test -p
hive-c0re -p hive-jobq` (321 + 40 passed) and `nix fmt`.
2026-08-01 16:06:06 +02:00
.forgejo/workflows docs: stop claiming tracker-tag/comment-block lint are non-required 2026-07-23 00:10:59 +02:00
branding docs(#1182): remove component-diagram.svg; trim README; link to website + options 2026-06-03 19:06:06 +02:00
claude-plugins claude-subagents skill: add concurrency guidance for container memory cap 2026-08-01 11:24:58 +02:00
docs docs(#2860): forge.url defaults to null, not a loopback address 2026-08-01 01:33:43 +02:00
frontend dashboard: consolidate NodeView.has_log into build_log_id 2026-08-01 11:38:23 +02:00
hive-agent stop telling agents to git against literal localhost:3000, point at $HIVE_FORGE_URL 2026-07-31 18:40:37 +02:00
hive-agent-mcp stop telling agents to git against literal localhost:3000, point at $HIVE_FORGE_URL 2026-07-31 18:40:37 +02:00
hive-agent-sock feat(#2635): wire harness-local questions mirror (inc2 pt2) 2026-07-26 02:07:33 +02:00
hive-bash-mcp hive-bash-mcp: flag bash-task completions with stderr instead of exit code 2026-07-31 15:48:50 +02:00
hive-c0re refactor(#2815): derive the transient pill from the running node 2026-08-01 16:06:06 +02:00
hive-core-agent-sock docs(#2627): add READMEs for the remaining infra crates 2026-07-23 13:16:29 +02:00
hive-forge hive-forge: attach the resolved repo to every verb's error 2026-07-29 21:18:32 +02:00
hive-forge-notify feat(#2642): a github.com notification poller alongside the forge one 2026-07-31 17:23:18 +02:00
hive-host-sock dashboard: consolidate NodeView.has_log into build_log_id 2026-08-01 11:38:23 +02:00
hive-jobq refactor(#2908): resolve a wire id in the graph, not the c0re layer 2026-08-01 15:53:46 +02:00
hive-matrix-mcp refactor(sock): one socket client, retry as a policy value 2026-07-26 22:44:48 +02:00
hive-metric docs(#2627): add READMEs for the remaining infra crates 2026-07-23 13:16:29 +02:00
hive-priv test(#2862): cover both fd/op mismatch branches in hive-priv 2026-07-31 22:15:37 +02:00
hive-priv-sock feat(#2862): receive a passed descriptor and stream a snapshot into it 2026-07-31 22:15:37 +02:00
hive-screen-mcp docs(#2627): add README for hive-screen-mcp 2026-07-23 14:17:47 +02:00
hive-sh4re remove request_next_turn: same-turn continuation is always worse than an external wake 2026-07-27 22:15:36 +02:00
hive-sock-client refactor(sock): one socket client, retry as a policy value 2026-07-26 22:44:48 +02:00
hive-types docs(#2627): add READMEs for the remaining infra crates 2026-07-23 13:16:29 +02:00
hivectl dashboard: consolidate NodeView.has_log into build_log_id 2026-08-01 11:38:23 +02:00
nix fix(#2898): carry hive/swarm display names as build-time options 2026-08-01 12:35:50 +02:00
scripts docs: stop claiming tracker-tag/comment-block lint are non-required 2026-07-23 00:10:59 +02:00
.gitignore fix(review): drop libnull.rlib artifact + add Errors doc to ensure_config_pr_webhook 2026-07-11 12:19:52 +02:00
.mailmap chore(#2165): add damocles@pr1ma + lexis@pr1ma mailmap entries 2026-07-04 13:50:16 +02:00
.prettierignore fix(#1997): exclude docs/tools/forge.md from prettier (list-item continuations) 2026-07-02 23:33:11 +02:00
.prettierrc temp: add prettier configs 2026-07-02 23:33:11 +02:00
Cargo.lock hive-c0re: wire up openapi spec + swagger ui (#2872) 2026-07-31 21:48:02 +02:00
Cargo.toml hive-c0re: wire up openapi spec + swagger ui (#2872) 2026-07-31 21:48:02 +02:00
CLAUDE.md docs(#2862): document the snapshot store, drop the dedicated option 2026-07-31 19:03:24 +02:00
clippy.toml hivectl: wireguard mesh setup verbs (#1756) 2026-06-19 14:37:50 +02:00
flake.lock nix flake update 2026-07-13 13:58:53 +02:00
flake.nix fix(#2860): make hyperhive.forge.url nullable instead of guessing a URL 2026-08-01 00:36:09 +02:00
README.md docs: shorten Overriding nixpkgs section 2026-07-16 00:06:17 +02:00

hyperhive

a swarm of claude-code agents, each in its own nspawn cage, gossiping over unix sockets. config changes flow as git commits, the operator approves them in a browser, every deploy is a tag. cyberpunk-themed dashboard included. 💜

Claude code is great in one window, exponentielle across many — but only if you can keep the agents from stepping on each other, give them durable identity, and stop them from eating production. hyperhive is the substrate.

  • identity = unix socket
  • communication = sqlite-backed broker (send / recv / ask / answer / remind)
  • config = git (manager proposes, operator approves, deploys land as tagged commits)
  • blast radius = container
host (NixOS, runs hive-c0re.service)
│
├── operator
│   ├── browser → :80 (hive-gateway)    dashboard + per-agent UIs
│   │                                   /agent/<name>/ → per-agent unix socket
│   └── CLI     → /run/hyperhive/host.sock   admin protocol
│
├── hive-c0re  (Rust daemon: lifecycle / broker / approvals /
│               auto-update / dashboard / sockets)
│
├── optional containers
│   ├── hive-gateway   nginx — proxies :80 → c0re dashboard + per-agent sockets
│   ├── hive-forge     Forgejo — per-agent accounts, config mirror (agent-configs/)
│   └── hive-matrix    tuwunel — Matrix homeserver + per-agent accounts
│
└── agent containers
    ├── h-ruth     manager (privileged MCP surface, approval gating)
    └── h-<name>   sub-agent (claude + MCP tools + per-agent web UI + unix socket)

→ website · → options reference

Depth lives in docs/ — pick the one matching your task:

reading path doc
dashboard layout + endpoints docs/web-ui.md (shape · dashboard · agent)
claude turn loop + MCP tools docs/turn-loop.md
config-edit + approval state machine docs/approvals.md
what survives destroy / purge / restart docs/persistence.md
naming, wire protocol, commit style docs/conventions.md
nginx vhost map + sub-domain routing docs/gateway.md
NixOS / nspawn gotchas docs/gotchas.md

Quick start

Minimal flake.nix for a host that runs hive-c0re:

{
  inputs = {
    nixpkgs.url = "github:NixOS/nixpkgs/nixos-26.05";
    hyperhive.url = "git+https://forge.darkest.space/hyperhive/hyperhive";
    # Pin hyperhive to your own nixpkgs instead of the one it ships with
    # (see "Overriding nixpkgs" below) — recommended for most hosts:
    hyperhive.inputs.nixpkgs.follows = "nixpkgs";
  };

  outputs = { nixpkgs, hyperhive, ... }: {
    nixosConfigurations.my-host = nixpkgs.lib.nixosSystem {
      system = "x86_64-linux";
      modules = [
        hyperhive.nixosModules.default  # hive-c0re + hive-forge + hive-gateway in one import
        ({ ... }: {
          services.hyperhive.enable = true;
          # services.hyperhive.c0re.operatorPronouns = "they/them";  # default: "she/her"

          # ... rest of your host config
          system.stateVersion = "25.11";
        })
      ];
    };
  };
}

hive-c0re opens its admin socket + dashboard, auto-creates the manager container, and auto-rebuilds any container whose hyperhive rev goes stale. claude-code is unfree — hyperhive scopes the whitelist to itself, nothing for the operator to set.

Overriding nixpkgs

hyperhive pins its own nixpkgs so it builds standalone in CI. Add hyperhive.inputs.nixpkgs.follows = "nixpkgs" (as in the quick-start above) to build it against your host's nixpkgs instead — one less nixpkgs evaluation, no version drift from the rest of your system. Standard flake follows pattern; works as long as your channel is reasonably close to the nixos-26.05 hyperhive develops against. Drop it again if a much older/newer channel hits breakage hyperhive's CI doesn't catch.

For the full list of host and agent NixOS options see the options reference.

Agent configuration

Per-agent config lives in each agent's agent.nix (proposed, operator-approved, deployed as git commits). Key options:

Multi-account Matrix support

hyperhive.matrixAccounts declares additional matrix accounts for an agent, beyond the hive-internal one. Each entry is keyed by account name and specifies:

  • tokenFile — path to the matrix bearer token (provisioned out-of-band)
  • sessionDir — path to the per-account matrix-sdk sqlite state (crypto keys + cache)
  • homeserver — optional homeserver URL (defaults to hyperhive.matrix.url)

Example:

hyperhive.matrixAccounts = {
  external-public = {
    tokenFile = "/agents/myagent/state/matrix-token-external";
    sessionDir = "/agents/myagent/state/matrix-sdk-state-external";
    homeserver = "https://matrix.org";
  };
};

The hive-internal account is always named main (synthesized from hyperhive.matrix.url + agent state). This option only declares extras; the main name is reserved and cannot be used here. Requires hyperhive.matrix.enable = true.

For more details see docs/matrix.md.

GitHub account

Every agent gets a managed GitHub identity — a gh CLI wrapper and git push over HTTPS — on by default (hyperhive.github.enable), inert until a PAT is provisioned. There is nothing per-agent to declare: paste an operator-supplied personal access token into the agent's dashboard credentials tab (or hivectl github set-token <agent> --token-stdin) and it works. The gh wrapper + git credential helper read the token live (git auths as x-access-token + the PAT; github.com only), so a rotated PAT takes effect with no rebuild.

Turn the integration off for the whole hive with the host option:

services.hyperhive.github.enable = false;

The PAT value is never in nix — only the enable flag. For more details see docs/github.md.

Operator CLI

hivectl is the operator-facing host CLI for ad-hoc administration that doesn't go through the broker (built alongside hive-c0re when the host module is enabled):

sudo hivectl forge create-user mara                       # provisions a forge user
sudo hivectl forge create-user mara --password 'hunter2'  # … with a fixed password
sudo hivectl matrix create-user mara                      # provisions a matrix user
sudo hivectl matrix create-user mara --password-stdin     # … reading one line from stdin

For agent names (i.e., a Coordinator::agent_state_root(name) exists), hivectl persists the resulting token to the agent's state dir like the boot sweep does. For non-agent names (e.g. the operator's own forge/matrix account), it prints the token to stdout and writes nothing.

Build / deploy

nix develop -c cargo check
nix flake check        # rust + nix + toml fmt + clippy

# deploy from a host config that imports hyperhive.nixosModules.default
nix flake update --update-input hyperhive
sudo nixos-rebuild switch --flake .#<host>