- Option doc: 'Default points at the in-host tuwunel (shared netns)'
→ updated to reflect that c0re forwards the gateway URL at runtime
- DNS oneshot comment: 'inert in shared-netns mode' → removed since
isolation is always on; oneshot always runs
Agents run in private netns (always-on isolation) and can't reach
host loopback directly. hive-c0re forwards HIVE_MATRIX_URL set to
`matrix.<domain>` via the gateway. Update two places that still
referenced the removed shared-netns path.
claude --effort now accepts low, medium, high, xhigh, max.
Expand EFFORT_LEVELS from 3 to 5, update NixOS enum + description,
and fix the stale doc comment in web_ui.rs.
Closes#2220
Root contexts (systemd services running as root, PID 1) default to
store=auto which resolves to the LOCAL nix store — bypassing the host
daemon, its remote builders (muede-pc2), and any prebuilt derivation
outputs already in the shared store. This causes spurious full
rebuilds of agent toplevels that the host already built and cached.
Two changes:
harness-base.nix:
- Add systemd.globalEnvironment.NIX_REMOTE = "daemon" — sets
DefaultEnvironment in systemd.conf so every unit in the container
inherits NIX_REMOTE=daemon. Non-root contexts already default to
the daemon socket; this only matters for root services that would
otherwise use the local store.
- Add NIX_REMOTE = "daemon" to environment.variables so interactive
shells also have it set (redundant with /etc/profile.d/nix-daemon.sh
but explicit and profile-agnostic).
hive-c0re.nix (hive-priv service):
- Add NIX_REMOTE = "daemon" to the service environment. hive-priv
runs as root and invokes nixos-container update + nix prebuild; these
must route through the host daemon so they see the shared store and
remote builders, not a private local store.
The sandbox-fallback = true in harness-base.nix is kept as a belt-
and-suspenders fallback but becomes a no-op for the common case once
nix routes through the daemon (the daemon builds on the host where
sandboxing works).
CLAUDE_CODE_OTEL_DIAG_STDERR was always set when OTEL is enabled, making
OTEL SDK errors (e.g. 404 from a misconfigured collector endpoint) appear
in every agent's stderr unconditionally. Move it behind a new opt-in flag.
Changes:
- nix/modules/hive-c0re.nix: add services.hyperhive.otel.debug (bool,
default false); wire to HYPERHIVE_OTEL_DEBUG env on hive-c0re unit.
- hive-c0re/src/meta.rs: add debug field to OtelConfig; read
HYPERHIVE_OTEL_DEBUG; emit hyperhive.otel.debug = true when set.
- nix/templates/harness-base.nix: add hyperhive.otel.debug internal
option; move CLAUDE_CODE_OTEL_DIAG_STDERR out of otelSettingsEnv into
a debug-gated lib.optionalAttrs block.
Default behaviour: OTEL exports silently (no stderr noise). Operators
troubleshooting collector connectivity set services.hyperhive.otel.debug
= true to re-enable the diagnostic output.
Add `baseClaudeEnv` (always-on) and extend `otelSettingsEnv` with the
env vars from #2149:
Always-on (every agent):
- CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC=1
- CLAUDE_CODE_PLUGIN_PREFER_HTTPS=1
- CLAUDE_CODE_RESUME_INTERRUPTED_TURN=1 (recovers MCP-flap mid-turn)
- CLAUDE_CODE_SIMPLE_SYSTEM_PROMPT=1
- CLAUDE_CODE_SYNC_PLUGIN_INSTALL=1
- CLAUDE_REMOTE_CONTROL_SESSION_NAME_PREFIX=hive+<agent>
- DISABLE_AUTOUPDATER=1 / DISABLE_UPDATES=1 (nix owns packages)
- DISABLE_INSTALL_GITHUB_APP_COMMAND=1
- DO_NOT_TRACK=1
- ENABLE_CLAUDEAI_MCP_SERVERS=0 (hive supplies its own)
- FORCE_AUTOUPDATE_PLUGINS=1
OTEL-gated (when hyperhive.otel.enable):
- CLAUDE_CODE_ENABLE_FEEDBACK_SURVEY_FOR_OTEL=1
- CLAUDE_CODE_OTEL_DIAG_STDERR=1
- OTEL_METRICS_INCLUDE_VERSION=1
Because baseClaudeEnv includes per-agent values
(CLAUDE_REMOTE_CONTROL_SESSION_NAME_PREFIX = "hive+${userName}"),
jq is now always run at build time — the OTEL-off branch that returned
the static asset verbatim is removed.
Per mara: a secret in the nix store is not acceptable. The non-secret
OTEL config (telemetry-enable, endpoint, protocol, resource attributes)
stays in the world-readable managed settings json; the auth header is
handled separately at runtime so it never touches the store.
New hive-otel-header oneshot (only when otel.enable && headersCredential
is set): inherits the forwarded otel-headers systemd credential via
LoadCredential, reads it at start, and merges OTEL_EXPORTER_OTLP_HEADERS
into the agent's 0600 ~/.claude/settings.json env block via jq. claude
layers the user env on top of the managed settings, so both the harness
turn-loop and hivectl choom (same agent user) export with auth. The
token is read from disk at start and never copied into the nix store or
the world-readable managed file.
Ordering is best-effort (before=, not a hard dep): a failure leaves the
harness running and telemetry exporting unauthenticated. headersCredential
option description updated to reflect it's now wired.
nix fmt clean.
The option description still claimed the credential is loaded via
systemd LoadCredential, but this PR removed that path. Clarify that the
option is currently inert (only the unauthenticated OTEL export is
implemented) and that runtime header injection is a planned follow-up,
so configuring it doesn't silently no-op without explanation.
Per mara: configure OTEL in the generated claude settings json (what the
Claude Code docs suggest), not a launch wrapper or /etc shell file.
claude-code auto-discovers /etc/claude-code/managed-settings.json in
every context — the harness turn loop AND hivectl choom — so putting the
OTEL env there gives telemetry parity declaratively, with no wrapper and
no --settings plumbing.
- managed-settings.json: was a static shared .source; now, when OTEL is
enabled, a per-agent build-time jq merge of the base asset + an env
block (jq at build, not eval-time readFile, to avoid IFD). OTEL off =
the static asset verbatim.
- otelSettingsEnv carries the static OTEL knobs + OTEL_RESOURCE_ATTRIBUTES
with the agent name (build-time) and the hive/swarm names forwarded by
meta.rs into environment.variables (mara: forward host config into
agent config where needed).
- removed the hive-serve-otel ExecStart wrapper, the per-unit otelEnv,
and the otel-headers LoadCredential from the harness service — the
harness binary emits no OTEL itself; only claude does, and it now reads
the settings json directly.
Known follow-ups (noted in code): the auth header (otel.headersCredential,
opt-in/default-null) is a secret and can't live in the world-readable
settings file — authenticated collectors need a runtime mechanism; this
PR covers the unauthenticated default.
nix fmt clean.
Follow-up to the harness fix: any unit whose ExecStart is a generated
wrapper (a NixOS `script =` block or a writeShellScript) gets its journal
SyslogIdentifier derived from the wrapper's store-path basename — an
opaque, unstable name. Pin an explicit stable identifier on every unit
hyperhive declares.
script/wrapper units (leaked store paths): forgejo-gpg-init,
hive-gateway-self-signed-cert, hive-tls-ca, hive-ci-prefetch,
hyperhive-isolated-dns, tea-login, forge-avatar-sync,
matrix-avatar-sync, hive-dashboard-links.
direct-binary units (basename already sane; explicit for tidiness +
future-proofing): hive-c0re, hive-priv, hive-matrix-daemon,
hive-bash-daemon, weston.
Only changes the displayed line identifier; journalctl -u <unit>
filtering (keys on _SYSTEMD_UNIT) is unaffected.
Under network isolation, agents lost DNS resolution of the hive's own
zones (forge.<domain>, matrix.<domain>): the hyperhive-isolated-dns
oneshot points resolv.conf at the bridge resolver, but resolvconf
regenerated it from host-tracking right after (dhcpcd re-triggering that
when the veth comes up), reverting to the non-authoritative host
resolver. Agents then couldn't reach the in-hive forge or matrix through
the gateway, which broke an isolate-by-default rollout.
Take resolvconf and dhcpcd out of the resolv.conf loop: disable
resolvconf and tell dhcpcd not to touch resolv.conf (without disabling
dhcpcd itself, so the veth still gets its address). Then the last writer
owns it -- the nixos-container host-copy in shared netns, or the oneshot
in isolated mode. Mirrors the approach the matrix container already
takes. Eval-checked; flake check passes.
Follow-up to the isolated-container egress fix. The hyperhive-isolated-dns
oneshot (which rewrites resolv.conf to the bridge resolver) was ordered
before network-online.target + tea-login, but NOT relative to the harness
(hive-ag3nt) or the matrix daemon. hive-ag3nt only declares
`after network.target`, so on first boot its first-turn api.anthropic.com
lookup could race ahead of the resolv.conf rewrite and error (self-heals
next turn, but flaps the first one).
Add hive-ag3nt.service + hive-matrix-daemon.service to the oneshot's
`before` so DNS is fixed before any network consumer starts. The matrix
entry is a harmless no-op when matrix is disabled (unit absent).
Caught by damocles in review of the parent PR.
When isolateContainers=true, claude (and all egress) broke in every
container: agents came up with an IP but no way off the bridge subnet.
Two container-side gaps, both confirmed against nixpkgs
nixos-containers.nix:
1. No default route. hive-priv wrote HOST_ADDRESS= empty in the nspawn
conf. nixos-container's container-side setup only installs
`ip route add default via $HOST_ADDRESS` when HOST_ADDRESS is
non-empty, so the container had an address but no gateway -> nothing
off-subnet (incl. api.anthropic.com) was reachable. Fix: write
HOST_ADDRESS=<bridge-ip>. In bridge mode the host-side address/route
setup is skipped, so this only affects the container's default route.
2. No usable resolver. nixos-container copies the host's /etc/resolv.conf
into the container at every start; the host resolver (e.g. 127.0.0.53)
is unreachable from a private netns and isn't authoritative for the
hive's own zones. Fix: hive-priv drops a marker carrying the gateway
IP only when isolated, and a new harness-base oneshot
(hyperhive-isolated-dns) rewrites resolv.conf to point at the bridge
dnsmasq. Inert in shared-netns mode (no marker), so the shared
container toplevel does the right thing in both modes.
The gateway IP is the address part of HIVE_NETWORK_SUBNET (the bridge IP
verbatim, honouring a non-.1 operator override), via a new validated
bridge_gateway_ip() helper with unit tests.
Unblocks defaulting isolation on.
Match the harness reader landed in #1600 (configured_effort reads
HIVE_DEFAULT_EFFORT, mirroring HIVE_DEFAULT_MODEL). Was
HYPERHIVE_EFFORT_LEVEL in the first cut; damocles standardized the
name on the HIVE_DEFAULT_* convention.
Per-agent enum option (medium|high|xhigh, default medium) rendered into
the HYPERHIVE_EFFORT_LEVEL env var. The harness resolves effort as
operator-override-file -> this env -> medium and passes it to
claude --effort at turn launch (harness read+apply tracked separately).
Same declarative-default + runtime-override shape as hyperhive.model.
Umbrella: operator-controlled claude effort.
Adds a per-agent hyperhive.availableModels list option (default
[ haiku sonnet opus ]) rendered into the HIVE_AVAILABLE_MODELS env var
(comma-separated) so the per-agent web UI model quick-picker lists
exactly the configured models instead of a hardcoded set. Operators set
a shared default hive-wide or narrow it per-agent.
An assertion guards that hyperhive.model is present in the list so the
picker can always offer the model the agent is actually running.
