The two "refactor later" items from REVIEW-2026-07-30.md; they share the wire
types, so they land together.
A capability had four divergent wire shapes — the HTTP handler, the MCP
adapter, pkg/client, and Maven's vendored copy of it. There is now a single
definition in pkg/client, mapped from domain by internal/wire and used by the
HTTP list/create/get paths and all four MCP surfaces. It lives in pkg/client
rather than internal so external consumers need not vendor internal/domain,
and so producer and consumer are literally the same type.
The unified shape is a strict superset of all four predecessors; nothing was
dropped. It adds enabled and requires_confirmation to the list responses
(never omitempty — an absent bool reads as unknown, not false), capability_id
to the MCP and client shapes, and the timing/attribute/version fields
previously only on get-by-ID. target_types and the list itself now serialize
as [] rather than null.
Both `id` and `capability_id` are deliberately kept, carrying the same value.
Maven decodes `id`; the spec and the rest of the API say `capability_id`.
Bearer auth is already a breaking change for that consumer, and stacking a
second silent one is the wrong trade — the redundancy stays until every
consumer is confirmed on capability_id, then `id` goes in an announced
removal. A test pins this and says so.
GET /api/v1/executions?entity_id=&since=&limit= implements spec §4.5, which
the Command Center needs. `since` reuses the changes-feed cursor convention
rather than inventing a second paging idiom. That cursor is the row's implicit
SQLite rowid, which is safe only while nothing deletes executions and nothing
VACUUMs — both would renumber and silently invalidate outstanding cursors. If
retention is ever added, this must become an explicit monotonic column first;
the constraint is documented at the query site.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Uea55zaiWuEByEDC4UBSdd
Finding 5 and the "remove" list of REVIEW-2026-07-30.md.
The tests were prose, not verification. TestContract_ChangesSinceAlwaysZero
documented a bug commit 74b19e0 had already fixed and could only t.Logf.
TestContract_TimeoutGoroutineNotCancelled was a bare t.Log, and is now
falsified by the preceding commit anyway.
TestContract_DestructiveCapabilityDisabledByDefault re-implemented
`risk != "destructive"` inside the test and asserted on its own local
variable — it would have passed if the handler were deleted. The real
derivation is now domain.EnabledForRisk and is tested against production code.
Also removed: the systemd provider (no systemctl in the distroless image and
nothing ever registered it), Server.ListenUnix, Registry.List,
DiscoveredTools, Capability.IsDestructive, the ExecutionDenied status,
EventCapabilityUnavailable, ResolveRequest/ResolveResult (superseded by
nexusclient), Engine.emitEvent, and eight unused Err values.
Three items on the review's list were kept, having turned out to be wrong:
ifString is still used by BuildCapabilities; EventExecutionDenied is asserted
on by a real test; and schema_migrations was NOT dropped. migrate() applies
migrations by slice index and writes user_version = index + 1, so removing an
element renumbers every later migration and any database past that point would
permanently skip one it had not yet applied. The live database is far behind
HEAD, so that is a data hazard rather than a cleanup. The CREATE TABLE is now
a no-op holding its slot, with a comment saying why the slot must stay.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Uea55zaiWuEByEDC4UBSdd
Finding 3 of REVIEW-2026-07-30.md. target_entity_id was accepted as any
non-empty string; the engine only compared it against a pinned TargetEntityID,
which is empty for every registered capability. Spec §4.3: "Hexis never
accepts a free-text target. Ever."
Targets are now checked in order: ent_ shape (free, never touches the
network), pinned target, existence in Nexus, entity still active, and a match
against the capability's TargetTypes. Validation runs before a confirmation is
consumed, so a bad target cannot burn one, and at confirmation-mint time too,
since a confirmation binds a target.
Two deliberate calls:
Nexus unreachable fails closed (503, ErrTargetUnverifiable). Failing open
would reinstate exactly this hole the moment Nexus blips, and hand it to
anyone able to degrade Nexus. Hexis holds no entity table, so "unreachable"
and "I cannot tell if this target is real" are the same statement. The cost is
that executes now require Nexus liveness; the lookup is bounded at 5s so a
hung Nexus fails fast rather than consuming the capability timeout.
An empty TargetTypes means no type constraint, not a bypass — the entity must
still exist, be canonical and be active. Rejecting empty outright would
disable 16 of the 19 registered capabilities, since only the docker.* entries
declare a target type.
The spec's stronger blessing guard is not implementable: Nexus has no blessing
concept at all. This is the achievable guard, and strictly weaker.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Uea55zaiWuEByEDC4UBSdd
Findings 1 and 2 of REVIEW-2026-07-30.md, which must land together: every
workspace capability registered with enabled=false, so the only working
provider could never execute. Fixing that alone would have turned a dead
execution path into a reachable one on an unauthenticated port.
Auth: a shared bearer token (HEXIS_API_TOKEN) is now required on the whole
/api/v1/ surface, compared with crypto/subtle.ConstantTimeCompare. /health
and /ready stay open for probes. It fails closed twice over — hexisd refuses
to start with an empty token, and the middleware returns 503 rather than ever
serving unauthenticated.
Guards: `enabled` and `requires_confirmation` are no longer readable from the
request body at all. Previously the handler derived the correct §4.3 default
and then let the caller override it, which is worse than no guard because it
reads as enforced. Both are now derived from the risk tier by shared helpers
in domain, used by the HTTP and provider registration paths alike;
unrecognised tiers fail closed to requiring confirmation.
BuildCapabilities sets Enabled, RequiresConfirmation and TimeoutSeconds
explicitly, and hexisd reconciles drifted rows on startup instead of skipping
any capability whose ID already exists — without that, allowlist edits never
reach an existing database.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Uea55zaiWuEByEDC4UBSdd
resolve_target previously returned a hardcoded "requires_nexus_resolution"
placeholder; it now calls Nexus's /api/v1/resolve via a new minimal
internal/nexusclient, configurable with -nexus (default localhost:8987).
/api/v1/changes ignored the since query param and always returned from
sequence 0 (`since = 0` regardless of what was parsed) — fixed to actually
parse and use it, so change-cursor polling works.
The MCP hexis.execute tool only forwarded capability_id/target_entity_id/
arguments/idempotency_key, silently dropping entity_version, requested_by,
origin, correlation_id, causation_id, resolution_evidence, and
confirmation_id even though the native HTTP API and domain.ExecuteRequest
already supported all of them — MCP callers now get full parity.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_018ghELqYhZNLub2TXGMazqA
Execution events (started/failed/succeeded) never carried the
correlation/causation IDs from the ExecuteRequest, even though the
fields existed on Execution. Added CausationID to ExecuteRequest and
Execution, and split emitEvent into a correlated variant used
throughout the execution lifecycle. handleExecute falls back to
X-Correlation-ID/X-Causation-ID headers when the body omits them.
Also add X-Hexis-Version negotiation on /api/v1/*, matching Nexus.
Part of Vikunja #273.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_018ghELqYhZNLub2TXGMazqA
Closes the biggest gap between the running execution engine and
ECOSYSTEM-SPEC.md §4.3: confirmations were entirely unmodeled, so any
capability could execute unconfirmed regardless of requires_confirmation.
- New confirmations table + Confirmation domain type; POST
/api/v1/confirmations mints a TTL-bound (120s) confirmation binding
capability id+version, target entity, and a sorted-key args hash.
- Execute() now requires a valid pending confirmation when the
capability demands one: rejects missing, expired, consumed, or
args/version-mismatched confirmations; consumes on success.
- Capabilities gain enabled (destructive risk defaults to disabled,
matching "must be turned on explicitly") and timeout_seconds.
- One in-flight execution per (capability_id, target_entity_id); a
second concurrent attempt is rejected (surfaced as 409 over HTTP).
- Wall-clock timeout per capability now wraps the provider call; on
timeout the outcome is "unknown" (new ExecutionStatus), never
"failed", and the run is never auto-retried.
- 9 new engine tests cover each guard from the spec's Phase 6 gate.
Vikunja #274.
Go daemon (hexisd/hexisctl) implementing capability registry, guarded
execution (confirmations, blessed-entity checks), systemd/workspace-mcp
providers per ECOSYSTEM-SPEC.md. Snapshotting existing working state
before further development.
kami