Watch

History

Repository files (latest commit first)
Filename	Latest commit message	Latest commit date
atlas 0a80f21003 prompts+docs: the lifecycle tools reach the whole subtree, not just direct children `require_descendant` (`socket_server/mod.rs:666`) authorises kill/start/restart/update/get_logs with `topology::is_descendant_of` — the caller's whole subtree, itself included. That has been true since ``53b4e752`` (#1865), whose message says "a parent owns its whole subtree; the root covers every agent as a consequence, no positional privilege", and two tests pin it (`is_descendant_of_in_grandchild`, `is_descendant_of_in_self_is_true`). The prose never followed. The four lifecycle tool descriptions, their `// IMPORTANT:` comments, `docs/tools/lifecycle.md`, the tools README, hive-agent-mcp's README and the system prompt every agent is rendered from all still said "direct children only" — while `list_containers`, four tools away in the same file, said "direct children + their subtrees". `lifecycle.md` also taught the model #1865 deleted: "Privileged agents (for example ruth) may operate on any sub-agent — the topology scope applies to all others." There is no privileged class to belong to; ruth reaches every agent because the check is transitive and everything sits under it. Same drift on the state-query side: `resolve_agent_state_target` is subtree-scoped by the same commit, so `get_loose_ends`' argument doc, the `QueryAgentState` capability doc and `docs/turn-loop/mcp.md` were all telling a parent it needs a capability to read a grandchild's threads. Two smaller corrections found on the way: * `list_containers` returns the caller itself. `is_descendant_of` is true for `candidate == ancestor` and `handle_list_descendants` filters the topology with it; called from a leaf agent it answers one row, that agent. * `request_init_config` accepts any unused name — the requester becomes its parent — or an existing agent already in the caller's subtree, not "a direct child". The editing surface is narrower than the guard, though: only direct children's config repos are bind-mounted, so re-seeding deeper in the subtree leaves no local copy to edit. `lifecycle.md` now says so. The prompt's other stale claim, the dead `request_apply_commit`, is #4226 and was fixed independently by damocles in #4227 while this was being gated. This branch keeps only the scope wording on that line. Closes #4225.		2026-09-11 15:23:50 +02:00
..
src	prompts+docs: the lifecycle tools reach the whole subtree, not just direct children	2026-09-11 15:23:50 +02:00
Cargo.toml	hive-c0re: deliver an agent's credential from the store to its state dir	2026-09-03 00:29:52 +02:00
README.md	docs: restructure into topic subdirectories, collapse duplicated index	2026-09-02 01:55:37 +02:00

README.md

hive-c0re

The unprivileged host daemon (runs as hive-core). Owns the sqlite broker, the approval/reminder/schedule queues, the generic job-DAG queue, container lifecycle, gateway/forge/matrix provisioning, per-container stats, and the axum operator dashboard. Largest crate in the workspace — bin-only, no separate lib.

When to use it

Host-level, cross-container orchestration: spawning/rebuilding/ destroying agent containers, the approval flow, dashboard-visible state, provisioning per-agent forge/matrix/gateway accounts. Agent-side behavior (turn loop, MCP tools) lives in hive-agent/hive-agent-mcp instead — this daemon only talks to agents over the socket wire types in hive-sh4re.

Shape

Cohesive clusters live in directory submodules, each re-exported at the crate root (crate::broker::… keeps resolving regardless of which subdirectory a module actually lives in). One line each — read the module's own //! doc-comment for real detail, don't expect this file to track it:

dashboard/ — the operator dashboard (containers, approvals, schedules, logs, topology).
job_queue/ — the job-DAG queue + desired-state reconciliation (docs/scheduler/coordinator.md).
lifecycle/ — nixos-container lifecycle + per-agent config flake generation.
stores/ — sqlite-backed stores (broker, queues, audit, power).
workers/ — background sweeps (crash watch, scheduled prompts, auto-update, knowledge sync).
agent_config/ — per-agent registries (tool groups, capabilities, resource limits, topology).
stats/ — dashboard metrics aggregation + OTEL export.
socket_server/ — the unix-socket request server shared by per-agent + manager sockets.
forge/ — optional Forgejo wiring (docs/integrations/forge.md).
coordinator.rs — top-level wiring for serve.
meta.rs, migrate.rs — the meta flake + schema/state migrations.
matrix.rs, gateway_nginx.rs, webhook_secret.rs, priv_client.rs — matrix provisioning, gateway vhosts, webhook secrets, and the hive-priv client respectively.

See the top-level CLAUDE.md/docs/ index for the full reading-path map.