Phase 6 of the roadmap. Populates the identity/teams and
orchestration/hierarchy module directories (previously empty or
near-empty scaffold stubs) and replaces identity/agents' pre-existing
placeholder, which queried a generic User/UserRole model unrelated to
this system's real architecture (nothing since 002-saas-integration's
CustomerReference has used it).
Teams and agents: active/inactive CRUD, never hard deletion; team
deactivation never cascades to its agents. Skills: compound-unique
upsert on (agentId, skillTag), never duplicates. Availability: a single
current record per agent, deliberately last-write-wins rather than
optimistic-locked — operational telemetry, not a durable business
record. The dynamic hierarchy: a nestable, orderable HierarchyNode tree
with every field (scope, assignment strategy reference, entry/exit
conditions) stored as opaque admin-set data; cycle detection runs only
on reparenting edits (a new node can't form a cycle); every
create/edit/activate/deactivate is audited via AuditLog, reusing
002's existing writer pattern. A capability-eligibility read path
composes hierarchy scope with caller-supplied skills for the future
orchestration/assignment phase (Phase 7) to call, deliberately
excluding availability per doc 05's own capability-before-availability
ordering.
Found and fixed a real bug before it reached tests: the initial
availability upsert reset currentLoad to 0 on every update, not just
creation.
Adds 11 unit tests (cycle detection, capability matching) and 4
integration test files covering all four user stories. Full regression
(every pre-existing 002-005 integration test plus all new ones) run
against real Postgres/Redis/MinIO: 107 passed, 9 skipped (005's
AI-key-gated tests, unrelated), 0 failed.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
33 tasks across setup, foundational schema work, and four user stories
(teams/agents, skills/availability, the dynamic hierarchy tree, and a
capability-eligibility read path for the future orchestration phase).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Reuses the existing identity/teams, identity/agents, and
orchestration/hierarchy scaffold directories per doc 07's documented
module placement. Key decisions: last-write-wins availability (not
optimistic locking — operational telemetry, not a durable record),
compound-unique skill upsert, cycle detection only on reparenting edits
(not creation, which can't form a cycle), and a capability-eligibility
read path that composes hierarchy scope with caller-supplied skills
while deliberately excluding availability per doc 05's own ordering.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Phase 6 of the roadmap: Team/Agent/AgentSkill/AgentAvailability models,
dynamic HierarchyNode configuration, and a capability-eligibility read
path for the future orchestration phase to call. Supersedes the
pre-existing identity/agents scaffold stub (which queried a generic
User/UserRole model unrelated to this system's real architecture);
leaves identity/customers and identity/auth untouched as out of scope.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
