classifyAndGate's decision pipeline moves to core: PolicyService runs the
full gate order (classify + transport escalation, plan-first, syntax,
host-only/host-lxc, VM QGA preflight, dedup, approval-flood, window
routing) over ports.GovernanceStore; ExecutionService records and
dispatches (auto-run via ssh.CommandExecutor + TargetResolver, queue via
ExecutionRecorder) with one converged path for run/docker_exec. Gating
matrix test added (risk x window x declared risk -> outcome); pair
coverage 95.6%.
Bug fix surfaced by the matrix: the flag-space syntax regex was inverted
— it refused valid 'tail -n 3' and missed the actual 'head - n' typo.
Fixed to match dash-space-value only.
Remaining Phase 4 items tracked in the plan: ApprovalService.Decide
convergence, execlog fold, execworker poller. VERSION 0.35.0.
Problem: relationship create/end existed as three drifted copies
(HTTP CreateRelationship/EndRelationship, MCP create_relationship/
end_relationship) with inline SQL, no ontology edge validation on
either path, and no audit on the MCP path.
Change:
- Internal/adapters/postgres/repositories.go: RelRepo implements
ports.RelationshipRepository (Create/End/ListFor) over the pool,
with in-tx upsert + audit/event side effects on Create.
- Internal/core/app/relationships.go: RelationshipService validates
edges against the cached ontology TypeTree (tree.ValidateEdge) and
delegates the tx to the repository. The adapter resolves slug→entity
and extracts types before calling the service.
- HTTP CreateRelationship: resolves source/target via ReadModels,
passes resolved types to RelationshipService for edge validation.
EndRelationship calls the service directly (audit stays in the
adapter for End — a simple toggle with no ontology check).
- MCP create_relationship/end_relationship: rewired to the service
(pool resolves entity IDs inline for the tool handlers; the service
validates edges and writes audit). The MCP path now gets ontology
validation and audit coverage for the first time.
- Composition root: RelationshipService built with RelRepo + Ontology
and wired through httpapi.NewHandler, ListenAndServe, and MCP
constructors.
Verification: go build/vet, full test suite (19 pkgs, DB integration
postgres+mcp green).
Problem: entity mutations (create/update/state) existed as three drifted
copies — HTTP CreateEntity/PatchEntity, MCP create_entity/
update_entity_attributes/set_entity_state — each with its own inline
SQL, its own validation subset (MCP validated lifecycle states, HTTP
did not; HTTP patched attributes without regenerating derived checks,
MCP did; MCP wrote no audit trail), the exact drift ADR 0016's first
vertical slice exists to collapse.
Change:
- internal/core/ports: DerivedCheck, Idempotency (adapter-owned request
hash + cached-body renderer so the replay record commits in the
create's transaction), IdempotentResponse + GetIdempotent read,
AuditEntry gains Method/Path/CorrelationID, Event gains
CorrelationID; EntityUpdateInput carries ExpectedVersion +
RederiveChecks (derivation for updates runs repo-side: the graph
host fallback reads relationships through the open tx).
- internal/adapters/postgres/repositories.go: EntityRepo (Create/
Update/SetState/reads/idempotency) preserving the load-bearing
check-then-act invariants in-tx: version WHERE-clause, declared
transitions + preconditions (ValidateTransition), duplicate-slug
mapping, audit/event/writeCheck all inside one BEGIN…COMMIT.
OntologyRepo: TTL-cached OntologyStore.
- internal/core/app/entities.go: EntityService — ontology validation
(type exists, concrete, state declared — the stricter MCP rule now
governs both surfaces), default-state resolution, id generation,
derivation for creates, audit/event construction, idempotency
pass-through.
- httpapi CreateEntity/PatchEntity rewired to the service; PATCH now
regenerates derived checks (the A2 parity gap). MCP create/update/
set-state tools call the same service — and now write audit + event
rows like the HTTP surface always did.
- Integration-test seed paths fixed for the adapters/postgres package
depth (../../../seeds).
Pre-existing failures documented: TestAPIEndToEnd (entity_types 60 vs
59; 501-endpoint now 200), TestClientLifecycleEndToEnd, TestPhase3*
rows — verified failing identically at ec11956 (scratch approval-
notifier commit test drift), unrelated to this change. All mutation
integration tests (create/patch/idempotency/audit/regeneration) pass.
Verification: make test-db (postgres + mcp green, httpapi green except
the pre-existing set), full non-DB suite (19 pkgs), golangci on new
packages — 0 issues.
Problem: check derivation logic lived in internal/checkdefaults with
the pure decision logic (buildKind, address/user/port resolution)
interleaved with tx I/O (entity_status insert, graph host fallback,
check upserts) — and internal/db importing it was the plan's called-out
inverted dependency.
Change:
- internal/core/app/checkdefaults.go: Derive(tree, target, lookup) —
the full derivation (monitoring overrides, host fallback via an
injected HostLookup thunk, per-kind builders) with zero I/O imports.
Types renamed for the app surface: CheckTarget, CheckDef,
DeriveResult, Skip; LogDeriveResult.
- internal/adapters/postgres/checks.go absorbs the I/O half:
EnsureChecks (entity_status row + upsert loop), writeCheck, and
hostViaGraph. The db→checkdefaults edge is gone — adapters→core is
the ADR 0016 direction (the Phase 7 SeedService note anticipated
this; the inversion is fixed a phase early).
- seed.go pending-checks loop uses app.CheckTarget + EnsureChecks;
mcp formatting/tests follow the renamed types; both test files moved
to internal/core/app.
- Deliberate behavior note: a hostViaGraph read failure inside the
thunk now logs a warning and degrades to 'skipped: no address'
instead of aborting the whole entity-create tx — a monitoring
derivation gap is visible (warn log + coverage sweep) and self-heals
on the next mutation; failing the create over a graph-read blip was
disproportionate.
Verification: go build/vet, full test suite green (app tests exercise
every buildKind branch at their new home).
Problem: the hexagon's Phase 2 (plans/2026-08-15-hexagonal-architecture.md)
must give the use-cases-to-be their contract surface: driven-port
interfaces, test fakes, the secrets interface moved into core, and the
postgres package inside the adapters tree — before the first vertical
slice (Phase 3) can wire a composition root.
Change:
- internal/core/ports: full driven-port catalog per plan §3.3 —
repositories as transaction-scoped aggregates whose inputs carry
derived checks, audit, and events (§3.6), plus CommandExecutor,
TargetResolver, Checker, Secrets, EventPublisher, Provisioner.
Port-local payload types (Event, AuditEntry, CheckDef, KnowledgeEntry,
ExecResult) keep signatures off infrastructure; TypeTree aliases
internal/ontology (pure over domain) until checkdefaults is absorbed.
ReadModels intentionally not declared yet — it materializes with the
Phase 3 slice and grows as report handlers rewire.
- secrets.Backend is now an alias of ports.Secrets; implementations
(Infisical, SOPS, Manager) unchanged. mcp's local secretBackend
subset is deleted; tool constructors take ports.Secrets.
- internal/db → internal/adapters/postgres (mechanical import rewrite;
package identifier stays db until the Phase 3 repository split).
sqlc.yaml, Makefile, golangci exclusions, and docs follow the move;
make generate-check verified.
- internal/adapters/ssh: Executor implements ports.CommandExecutor over
the actuator dial pool + RunStreaming (10-min default timeout carried
over from the httpapi path).
- internal/adapters/remote: Resolver implements ports.TargetResolver
delegating to internal/remote (still pool-based; drops onto
ports.EntityRepository when repositories land in Phase 3 — documented
transitional import).
- internal/core/ports/portstest: importable fakes — in-memory
EntityRepo (with check-then-act SetState, side-effect recording),
RecordingExecutor, FakeChecker, SpyPublisher; port-satisfaction
guards; tests.
Risk: ports are declared ahead of implementations — signatures firm up
per phase as slices land (documented in the package doc); the
remote→postgres transitional import is explicit and dissolves in
Phase 3.
Verification: go vet, make test (race, 19 packages), generate-check,
golangci on core+adapters — 0 issues; full-repo baseline down
365→344.
Removes the entire Matrix-based notifier (internal/notifier/) that polled
for pending approvals, sent Matrix alerts, and checked for reaction-based
approve/deny. Approval decisions now work on any chat platform (Hermes
desktop, Telegram, Discord, WhatsApp, CLI) via two new MCP tools:
- list_approvals — query pending/recent approvals by status or entity
- decide_approval — approve/deny via same API endpoint as UI + nomos
Config fields removed: MatrixHomeserver, MatrixUserID, MatrixToken,
MatrixRoomID, ApprovalHMACSecret. Docker notifier: service removed.
Approval HMAC token generation removed (unused by code).
The existing chat-assent path in nomos (cmd/nomos/assent.go) and the
control-room Approve button keep working unchanged — both call the
shared POST /api/v1/approvals/{id}/decision endpoint.
New tool batches session discoveries into the knowledge graph:
- Creates a session-audit knowledge entry with summary
- Links it to all touched entities via 'documents' relationships
- Creates individual discovery knowledge entries
- Stamps each entity with last_agent_session attribute
- Updates AGENTS.md with tool listing
P0: Transport escalation in classifyAndGate (server.go:745) now exempts
log-inspection commands (tail/head/cat/journalctl on *.log or /logs/)
from the /opt//etc//var/lib/ gating on LXC targets. Fixes the
'tail -3 /opt/seanime/data/logs/seanime.log queued for approval' bug.
P1: queryEntity returns actionable error when slug_or_id param is empty
('slug_or_id is required') instead of silent 'entity not found: '.
P2: Added slugArg() helper (server.go) so get_entity, get_relations,
get_blast_radius, and explain accept 'slug' as an alias for their
declared param name. Solves the discoverability inconsistency where
every tool used a different param name for the same concept.
P3: Two new MCP tools:
- restart_service(target, service) — systemctl restart wrapper,
correctly classified config_mutation (requires approval)
- push_file(target, source_path, dest_path, backup=true) — pct push
from Proxmox host into LXC, with optional backup. Classified
config_mutation. LXC-only for now.
P4: Updated homelab-lxc-ops skill with MCP tools preference table.
Plus: Wails desktop build now uses build-tag approach for frontend embed
(assets_embed.go + assets_stub.go), so go build ./... works on
clean checkout without the frontend built first.
Version: 0.31.0 → 0.32.0
Every run call creates a classification entity, but it was never
connected to the execution via a graph edge — only via a DB column
(executions.classification_id). The ontology requires:
classification —precedes→ execution
Without this edge, all 53 classification entities had zero
relationships, making them invisible to get_relations and
blast-radius analysis.
Adds an idempotent INSERT into relationships after the existing
classification_id update, matching the same pattern used for
targets edges.
- Wire secretsManager in NewHandler() — instantiate InfisicalBackend
when OIKOS_INFISICAL_SITE_URL is set (previously always nil)
- Add get_secret, list_secrets, set_secret MCP tools with nil-backend
graceful degradation
- Add oikos secret get|set|list CLI subcommands for Infisical
- Fix Set() bug: create-before-update so new keys are created;
add Type: "shared" to Update so it finds the right secret;
disable SDK cache so Get returns fresh data after Set
- Clean enrollment response: remove fake infisical_client_id/
infisical_client_secret stubs, store age key in Infisical for real
- classifyAndGate: escalate read-only commands on lxc: targets that
touch /opt/, /etc/, /var/lib/ to config_mutation. The classifier
scores command text only, not the SSH transport layer — SSH-ing into
a container to read config is riskier than pct exec from the host.
- ontology: standalone-server monitoring override from inherited
[ping, resource, updates] to [http]. VPS-like machines may not be
SSH/ICMP-reachable from the scheduler; HTTP is the LCD liveness
signal. Entities with full SSH can override per-entity.
- get_health_summary: filter out state=destroyed entities (was noise
from 20+ destroyed test LXCs, deprecated services, etc.)
- create_entity: document the monitoring footgun in the tool description
(creating a type=check entity does NOT wire a check_def; the correct
path is update_entity_attributes with monitoring + url attributes)
I — run pre-flights QEMU guest agent before queueing VM execution
classifyAndGate now checks vm: targets for qemu_guest_agent attribute.
If not_running/missing, returns immediate error instead of queuing forever.
II — policy.yaml: documented host-mutation classifier rule
Added comment clarifying that host-level package/kernel mutations
(apt-get install, dpkg, systemctl enable) always classify as
config_mutation and thus need operator approval.
III — health attribute read-only in update_entity_attributes
Strips scheduler-owned keys (health, last_check_at, last_check) from
attribute updates with a clear message directing agents to
get_health_summary / list_checks instead.
IV — Recorded discovered dependency edges
vm:zimaos → depends-on → lxc:nfs-export (NFS /media/library mount)
vm:zimaos → depends-on → host:strong (NFS /media/ludo-library mount)
Also updated the run tool description to mention both guardrails.
get_relations now accepts an optional 'types' (comma-separated) parameter
to filter relationship types — filters out the noisy exec/targets edges
that previously drowned useful host/provides edges.
get_health_summary now accepts an optional 'health' (comma-separated)
parameter to return only entities in specific health states (e.g.
'health=down,stale') instead of the full 100+ entity list.
ping_service now:
- Falls back to e.attributes->>'public_host' when 'url' is not set
(covers LXCs that only have public_host in the graph)
- Performs a live HTTP HEAD probe against the resolved URL, returning
the actual status code instead of just the scheduler's stale health
state
Also: fixed matrix.hubris.network DNS record (was pointing to dead VPS),
pruned 6 dead graph edges, wired url attributes on 7 LXCs, added VPS
HTTP monitoring check, and resolved the 18k-occurrence unmonitored signal.
This session's audit is documented as
document:nomos/2026-08-05-dns-monitoring-improvements-for-strong-hosted-services.
- Add MCP tool — lightweight connectivity check returning server
identity, no DB hit (resolves agent connection-test friction)
- Tighten 6 tool descriptions (get_relations, get_health_summary,
query_metrics, get_trend, get_event_timeline, ping) to be searchable
in the first 8-12 words
- Document Hermes MCP client setup in ADR-0012 with token security caveat
- Move completed plan to plans/done/
The DB-only audit_knowledge_graph can't see guests running in Proxmox that
have no entity, or entities whose pve_id is no longer live — the drift that
the stray test LXCs were a symptom of. discover_infra_drift enumerates running
guests via pct/qm list on every proxmox host (over the same SSH/pct path the
checks use) and diffs against the DB: returns missing (live, no entity) and
ghost (DB, not live). Read-only.
Companion to audit_knowledge_graph; the skill now runs both and treats the
remaining checks (misplaced parent, undeployed scripts, seed drift) as manual.
Adds audit_knowledge_graph (MCP tool) and GET /api/v1/audit/drift (endpoint)
backed by a shared internal/audit package. One pass surfaces the structural
gaps an operator otherwise finds by accident: orphan check entities, checks
targeting deprecated/destroyed entities, probes stuck down/unknown, unmonitored
declared types, and live edges pointing at destroyed targets. Each finding
carries a suggested remediation runbook. Read-only and safe to run unattended.
Ships the knowledge-graph-audit skill (SKILL.md + seeded runbook) that
interprets the report and routes findings to the lifecycle runbooks.
The scheduler SSHed each guest directly and assumed a deployed probe script
plus working root SSH at the guest's address — false for headless (nfs-export),
keyless (teddycloud), mesh-only (rclone), and macOS (mac-mini) targets, which
left 49 enabled checks stuck "down" on a healthy fleet.
Extract the MCP run tool's resolveExecTarget into a shared internal/remote
package and make it the single execution path for both the scheduler and MCP.
LXC/VM checks now host-hop via pct exec / qm guest exec through the owning
Proxmox host (no per-guest lan_ip, sshd, or authorized key needed); hosts and
workstations resolve their address and user live, so mac-mini's `user: dtoro`
is honored without a re-seed. Address preference now prefers public_ipv4 over
mesh, so netbird-vps is probeable from the scheduler container.
cpu_check.sh gains a real Darwin branch (it reported cpu_pct 0 before).
checkdefaults.resolveSSHUser reads the top-level `user` attribute too.
A machine-target resolution failure is now logged before falling back to baked
config, so a broken probe-config is distinguishable from a real outage.
Monitoring coverage was 3 of 89 active entities. Three bugs, each hidden by
discarded errors in checkdefaults:
- writeCheck generated a fresh uuid, inserted the check entity ON CONFLICT
(slug) DO NOTHING, then wrote a check_defs row referencing it. On any
re-seed the slug already existed, the entity insert no-oped, and the FK
violated — aborting the ingest transaction and surfacing as an unrelated
failure several entities later. Re-seeding has been broken since; prod's
coverage was frozen at its first successful seed. This is what
TestSeedIngestIdempotentAndNoDuplicateEdges had been reporting.
- shortSlug truncated to the last 8 chars, so all 21 ingress routes collapsed
to ".network" and overwrote each other; service:jellyfin collided with
lxc:jellyfin.
- The ssh-script checker never read the `args` config checkdefaults wrote, so
process_check.sh always ran without its unit name and returned "unknown".
Coverage is now 75/89. Monitoring is declared per entity type in
seeds/ontology.yaml and resolved through the is-a hierarchy, so a type can say
it warrants nothing (site, lan, mesh, cluster) and never be reported as a gap.
coverageSweep raises an `unmonitored` signal only where a type declares
monitoring it lacks — 8 real gaps, no false positives.
Also:
- entity_types.attribute_schema was never ingested: the seed loader read
"attribute_schema" but the YAML says "attributes", so all 60 types stored
JSON null.
- ListExecutions ignored its declared target/action/correlation_id filters and
paginated on a non-unique target slug, dropping and repeating rows.
- started_at was captured but only written at terminal state, so a running
execution reported NULL for its whole life. The three MCP auto-run copies
wrote no timing at all; they are now one autoRun helper.
- SSH output was buffered to completion and discarded entirely on timeout.
Both sshExec copies now stream through a shared execlog sink into
execution_logs, and keep partial output when a command is cancelled.
- executions.correlation_id was a random per-execution uuid that correlated
nothing; it is now the chat session id, which is what lets the chat tail
live output.
- reversible_low had no auto-run branch despite policy declaring it
unattended. Since computeCommandRisk never returns it, the class only arises
when an agent declares it over a read_only command — so gating it penalised
candor without adding safety.
- backup-target gains a backup-freshness checker (portable find -mmin, since
the first target is on macOS), resolving its host by walking backs-up-to
backwards. The pre-deploy pg_dump is now a tracked backup target.
UI: an Executions section on entity detail with live output tailing, and
streamed output under a running `run` call in the chat timeline.
Migrations 022-024. Ops.svelte and context.ts exclude execution.output from
their refetch triggers, which would otherwise fire once a second per command.
Co-Authored-By: Claude <noreply@anthropic.com>
A config_mutation/destructive run() queued for approval never touched
agent_sessions.status — only ask_operator did that, setting
awaiting_input. So a task blocked on an execution approval was
indistinguishable from one still genuinely working: the frontend's
"Needs input" bucket only checks status===awaiting_input (never lit
up for these), and the idle-sweep safety net only excludes
awaiting_input from its stale-task query, so after ~30 minutes idle
it would nudge the agent and then auto-close the task with
outcome=partial while the approval was still sitting there undecided.
classifyAndGate now flips the session into awaiting_input the moment
an execution is queued (internal/mcp/server.go), and DecideApproval
flips it back to executing once the approval is approved, denied, or
revoked (internal/httpapi/approvals.go) — mirroring askOperator /
answerQuestion's existing pattern for session_questions. Both emit
task.status so the board updates live.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Classifier now unwraps pct exec / qm guest exec / bash -c / sh -c / sudo
and env-var assignments before classification, so read-only inspection
wrapped in pct exec no longer escalates to config_mutation. curl GET
(default method, no -d/-F/-T/-o/>) is read-only. Eliminates the three
duplicate rclone sessions (a51e2086, 8acea2e3, cb8c8a4a) that bounced
off the classifier for the same goal.
New classify_command MCP tool: command-scoped preflight that returns the
exact risk class run would assign. Documented in SOUL.md with guidance
to pre-classify before run when the verdict is uncertain.
set_goal surfaces prior partial/failed sessions from the last 24h so the
agent picks up the thread instead of rediscovering it.
completeTask auto-closes in-flight plan steps (pending/running -> done
on success, skipped on partial/failure), so one-step plans no longer
need the per-step running->done dance right before completion.
Migration 021 adds blocker + closed_at to agent_sessions. completeTask
sets closed_at once and derives a structured blocker reason
(approval_timeout, user_abandoned, classifier_overreach, model_refusal,
tool_error, ...) from the last assistant message.
/sessions list now carries message_count, tool_call_count,
duration_seconds (server-side aggregates — no more N+1 transcript
fetches to audit a fleet). GET /sessions/{id} returns both metadata
and messages. New query params filter + paginate: outcome, status,
entity_id, blocker, since (RFC3339 or Go duration), cursor, limit.
Titles now prefer the goal when set; sessions without a goal fall back
to the first assistant text.
New GET /sessions/{id}/tool_calls flat view for audit scripts.
Plan: plans/2026-07-20-session-review-ten-sessions.md. VERSION 0.7.12 -> 0.7.13.
Session-review implementation for the three sessions audited in
plans/2026-07-18-session-review-three-sessions.md. v0.7.11 → v0.7.12.
P0.1 — retry cap + investigate-before-retry (cmd/nomos/retrycap.go,
agent.go): after 3 identical failing run calls in a single turn, refuse
to dispatch the call again and return a directive to investigate *why*
(ps/strace/lsof) or surface the blocker. Per-turn scope so a fresh turn
after the operator responds can retry once more. Session 1e9c7691's 20+
identical chown retries (knfsd held a kernel lock on the exported NFS
dir) is the direct motivation.
P0.2 + P1.8 + P2.10 — SOUL.md guidance: hung command is not a failed
command (investigate before retry); ask before proposing a multi-step
migration; multi-goal sessions summarize the arc not just the last goal.
P1.3 — two new runbook entities in seeds/knowledge.yaml:
- nfs-exported-dir-mutation-hang (the knfsd fchownat lock procedure:
killall → exportfs -u → mutate → exportfs -a → verify)
- netbird-mgmt-oidc-race-after-upgrade (docker restart netbird-mgmt
after ~30s for the traefik/authentik OIDC race)
P1.4 — setGoal emits task.superseded event when prior goal is overwritten
by a different goal (store.go, TestSetGoal_SupersededEvent). Session
55927f0a had two set_goal calls with the first silently abandoned.
P1.5 — inspect_path MCP tool: runs mount/df/ls/stat for one path across
up to 8 targets in one parallel call, replacing the 15+ run-call
fact-gathering fan-out sessions 1 and 2 each spent on cross-target path
tracing (tools.go, server.go: inspectPathAcrossTargets, inspectOneTarget).
P1.6 — vm: target support in run via qm guest exec (no more SSH-hop
with nested quoting). Extracted shared resolveProxmoxHostSlug for
LXC + VM, with hosts-relationship fallback when attributes.host is
absent (server.go, tools.go). Session 55927f0a's SSH-hop workarounds
for vm:zimaos are the direct motivation.
Deferred (documented in plan): P1.7 (approval window auto-extend on
timeout) and P2.9 (long-running command PENDING detection) — both
addressed at lower cost by the retry cap. Session 3's poll-after-timeout
pattern already works; the cap protects against the failure mode.
Deleted 8 genuinely unused sqlc queries (no inline equivalent):
- UpsertCurrentRelationship, ListEntitiesCapped, ListEntityStatus,
UpdateSignalState, InsertClassification, InsertFeedback, InsertSkill,
UpsertCurrentRelationship — all had zero call sites.
Migrated 9 inline raw SQL sites to use sqlc queries:
- GetOntology (impl.go): ListEntityTypes, ListRelationshipTypes,
ListLifecycleDefs — replaces 3 raw pool.Query blocks with typed sqlcgen
calls, eliminating manual row scanning.
- EndRelationship (phase3.go): EndCurrentRelationship — replaces tx.Exec
with sqlcgen.New(tx).EndCurrentRelationship.
- checkPrecondition (impl.go): GetEntityStatus — replaces tx.QueryRow +
manual Scan with sqlcgen.New(tx).GetEntityStatus.
- GetEntityRelations (impl.go): ListEntityRelations — replaces raw pool.Query
+ scanRelationships helper (now deleted).
- GetGraph (impl.go): ListGraphEdges — replaces raw pool.Query +
scanRelationships.
- resolveEntityID (impl.go): GetEntityBySlug/GetEntityByID — replaces
raw pool.QueryRow + Scan.
- createApproval (mcp/server.go): InsertApproval — replaces raw pool.Exec
with sqlcgen.InsertApproval.
Deleted scanRelationships helper (was only used by the two migrated
graph queries above).
Regenerated sqlcgen — also picks up stale model updates (AgentSession,
SessionPlanStep, SessionQuestion, etc. from recent migrations).
Documented the carve-out in .agents/dev/CONTRIBUTING.md §SQL conventions:
sqlc is the default; raw pool.Query/Exec is reserved for LISTEN/NOTIFY,
dynamic WHERE builders, blast_radius(), and COPY.
go vet, build, httpapi/mcp/db tests all pass. -383/+170 lines.
P1: add docker compose (logs|ps|top|config|images|port|cp) to read-only
allowlist. docker compose logs was classified as config_mutation, causing
individual approval cards for read-only inspection commands.
P2: remove approval entries from activityLog. They were always status=running
and never transitioned to done (the derived store builds from tool-call
text, not execution status), causing AgentIndicator to latch onto a stale
'Approval: ...' entry and never clear — even after the session completed.
P3: remove InlineApproval from Chat.svelte. The green 'Completed in 1s on
lxc:...' boxes were noise in the chat stream. Approval UX belongs in the
Operations page (already has it via Ops.svelte), not inline in the chat.
P4: stale execution cleanup. Startup sweep (mark >1hr non-terminal as
cancelled) + 5-min periodic sweep (mark >10min non-terminal as cancelled).
98 orphaned executions accumulated from eval testing (39 running from
apt_upgrade:audit timeouts, 19 pending_approval, 3 approved).
P5: refuse second config_mutation run when an approval is already pending
for the session. Without this, the agent queues N individual approvals
before the operator can respond — confirmed in session 20757eb9 (two
approval cards for what should have been one plan-level approval).
VERSION 0.7.0 → 0.7.1
P1 plan-first: run handler refuses without propose_plan (structural gate,
not SOUL.md prose). Plan window decoupled from set_goal — config_mutation
auto-run only on operator approval (assent window). Closes the approval-free
config_mutation hole confirmed in session d0d562e0.
P2 iteration: reopenSession flips terminal→executing, marks prior plan steps
replaced, clears outcome. proposePlan excludes replaced from in-flight check,
bumps generation. A follow-up on a completed session starts a new sub-task
with a fresh plan — no more errPlanInFlight dead end.
P3 reasoning: accumulate per-iteration text into the persisted row instead
of overwriting with the last text event. Reload shows intermediate thinking,
not just the final summary.
P4 read-only allowlist: add find, tree, locate, systemctl list-timers/
list-unit-files/show, timedatectl, hostnamectl, systemd-analyze, rclone
ls/lsl/md5sum/check/cryptcheck. Fixes the find misclassification from
d0d562e0.
P5 eval harness: new assertion kinds (proposes_plan, plan_before_run,
plan_generations), multi-turn followups, fetch /sessions/{id}/plan. Four
manifests under evals/.
P6 SOUL.md: strip degenerate-case carve-out, add ITERATE step, update
set_goal guidance.
VERSION 0.6.0 → 0.7.0
Ships the 9 remaining post-fix items and a golden-conversation eval harness
that validates them against the live agent. All 4 evals pass.
SOUL.md (F.1, C.2, E.1):
- Consolidated three overlapping task-flow sections (MANDATORY TASK FLOW,
'Every chat is a task', 'AFTER EVERY TASK: WRITE BACK') into one. ~50
lines shorter. The operator's 'be more crisp' feedback.
- Added anti-patterns: don't re-execute on UI/sidebar complaints (C.2);
don't re-run fleet-wide audits when same-day knowledge exists (E.1).
- Updated approval vocabulary in step 4 to match tasks.go (approved/yes/
go/proceed/continue/ok/go ahead).
Tool-result strings (F.2):
- set_goal: tightened to 'Goal set. NEXT: pre-plan (read-only tools only).
Then propose_plan. Do not call run.'
- update_plan_step: added '(Advance with update_plan_step + run; do not
re-propose.)'
C.1 — completeTask rejects re-completion of a terminal session:
- Returns errTaskAlreadyComplete when status is already done/failed.
- The tool result directs: 'Task is already complete. Do not call
complete_task again. If the operator pointed out a UI/sidebar
inconsistency, fix it with update_plan_step...'
B.4 — Surface real model error text:
- chatWith's error event now includes finish_reason + refusal text:
'Nomos returned an empty or unusable response (finish_reason=length).
Retry or rephrase.' instead of generic 'empty response'.
- The resume-failed note already carried errText (B.3), which now has
the real context.
B.5 — Back off between resume retries (4s, 8s):
- resumeSession now sleeps before attempts 1 and 2 (exponential backoff).
A transient provider issue gets time to clear instead of 3 identical
calls in 3 seconds.
B.6 — Don't persist the empty placeholder as a visible bubble:
- If a chat turn ends with no text and no tool calls (model empty-response'd
and all retries failed), delete the placeholder row instead of persisting
an empty bubble. The error was already streamed via done+error=true.
E.2 — list_lxcs last-audited hint:
- The list_lxcs result now includes last_audited_at — the most recent
knowledge entry (tagged audit/update, or titled audit/update) linked
via an 'about' edge. The agent can see 'nextcloud — last audited today'
and skip re-running it.
Tool-call doubling bug fix (found by the eval harness):
- main.go + continue.go: the tool_use and tool_result events were both
appending separate entries to the persisted tool_calls array, doubling
every tool call in the transcript. Confirmed pre-existing (d9cdcee1,
v0.3.x era). Fixed: tool_use creates the entry, tool_result merges the
result into the same entry (matched by id). One entry per tool call.
Golden eval harness (cmd/nomos/eval/):
- A standalone Go program that loads YAML manifests of golden conversations
+ assertions, sends prompts to the chat endpoint, drains the SSE stream
(keeping the agent's context alive), and scores structural assertions
against the persisted transcript.
- 4 golden conversations covering: trivial read-only (degenerate case),
plan + proceed (the original duplication bug), UI complaint (no re-exec),
fleet audit (knowledge preferred over re-execution).
- Structural assertions only (tool-call sequences, plan steps, writeback,
completion) — text quality is model-dependent and not scored.
- Run: go run ./cmd/nomos/eval -gateway http://localhost:8092 -manifest
cmd/nomos/eval/evals/*.yaml (~$0.10/run in OpenRouter credits).
Eval results (4/4 passed):
trivial_readonly: 2 tool calls, no plan, no run
plan_advances_on_proceed: 13 tool calls, propose_plan x1, writes back
ui_complaint_no_rerun: 12 tool calls, propose_plan x1, writes back
knowledge_preferred_over_rerun: 7 tool calls, search_knowledge x1, 0 run
Version 0.5.2 -> 0.5.3 (minor: eval harness + structural hardening).
SOUL.md: mandatory 6-step task flow at TOP of file, unmissable.
Agent MUST: set_goal → pre-plan (research only) → propose_plan → STOP
and wait for approval → execute (auto-run under plan window).
Backend:
- set_goal now opens plan window immediately (config_mutation auto-runs)
- set_goal result tells agent to do pre-plan + propose_plan, not run
- propose_plan result tells agent to STOP and wait for approval
- plan window value unified to 'active' (set_goal + propose_plan)
This prevents 23 individual approval popups — one plan approval instead.
Backend:
- proposePlan sets plan window in autonomy_settings (nomos:plan:<session>)
- run handler checks plan window — auto-executes config_mutation commands
within plan without per-action approval
- planWindowActive function in server.go
- Plan window cleaned up on completeTask (already covered by LIKE '%:' || )
Frontend:
- Removed 'Session graph' header bar
- Cooler empty states: Plan shows animated dots + 'Awaiting plan…',
Activity shows pulsing dots + 'Waiting for activity…'
12 of 33 MCP tools now render as rich inline cards instead of raw JSON:
EntityCard, HealthSummary, LXCList, EntityTable, KnowledgeResults,
BlastRadius, ChangeLog, FleetSnapshot, MetricChart.
Architecture:
- Server: annotateJSONResult() wraps queryRows with __renderer hints
- Registry: match/dispatch system maps tool names to Svelte components
- Chat: inline dispatch with 5-card limit, overflow to collapsed group
- ToolCallGroup: unmatched prop, hides when all matched, ARIA labels
Tests: 3 new Go tests for annotateJSONResult (wrap, no-op, multi-row).
search_knowledge and get_entity_knowledge only ever returned a ts_headline
snippet/short headline — enough to find a note, not enough to act on it.
Add get_knowledge_content(slug), mirroring the web UI's
/api/v1/knowledge/content/{id}, so the agent can read a document/
investigation/runbook's full markdown body once it knows which one it
needs. upsert_knowledge already covered the write side. Cross-referenced
all three tool descriptions so the agent discovers the full-read path.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Executions were being created with no outgoing edges to what they acted
on or which task/session drove them, silently starving the graph of new
data going forward — found during this session's DB audit, which had to
backfill 245+25 missing targets/involves edges for existing executions.
This closes the gap at the source: every execution now gets a
target-->targets-->execution edge, and (when the caller supplies a
session/task) a task-->involves-->execution edge, both idempotent
(NOT EXISTS guards) so retries/backfills don't duplicate.
Two call sites: the deduped systemctl/apt_upgrade/pct_create fast path
and the general classifyAndGate path.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Events, Agent, and Audit were standalone read-only pages that never
cross-referenced the entity they related to. Fold them into EntityDetail
as entity-scoped cards (Agent activity, Audit trail) alongside the
existing Signals/Executions/Knowledge cards, and give the Signals card
real Ack/Mute/Resolve actions. Signals stays a standalone page since
it's the only one with cross-entity triage value (badge count, actions).
Also fixes the underlying reason those new cards would've stayed empty:
agent_activity rows were never tagged with entity_id at insert time
(cmd/nomos/store.go, internal/mcp/server.go), even though the column
and the API filter both support it. Added a best-effort resolver that
checks common tool-arg keys (target, entity_slug, slug, ...) against
the entities table.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Fixes B1 and B2 of plans/2026-07-11-nomos-agent-code-review.md together,
since the right granularity for B1 in the auto-continuation worker turned
out to require B2's restructuring anyway (see below).
B1: grep -rn "recover()" cmd/nomos/ internal/mcp/ internal/httpapi/ returned
nothing before this — every explicitly-spawned goroutine (continuation
worker, resumed chat turns, async execution dispatch, the SSE listener, two
duplicate sshExec implementations' output-collector goroutines) crashed the
whole process on an unhandled panic, not just that one goroutine. More
consequential post-concurrency: more simultaneous unattended background work
means more surface area for one bad input to end every running task.
New internal/safego package: Go(label, fn) launches fn in a goroutine with a
recover-and-log wrapper. Applied at every bare `go` spawn site across the
three packages. Two sites needed bespoke handling instead of the generic
helper because their callers block on a channel and a silent recover would
just make them hang until timeout: sshExec's output-collector goroutine (two
near-identical copies, internal/mcp/server.go and internal/httpapi/phase3.go)
and httpapi's ListenAndServe goroutine — both now recover AND send a
synthetic error result so the waiting select unblocks immediately instead of
waiting out the full timeout.
httpapi's sseListener got extra treatment: its per-notification handling was
extracted into handleNotification with its own recover, so a panic decoding
ONE malformed pg_notify payload can't kill the listener goroutine for every
connected SSE client — the outer goroutine spawn only needs to guard the
connection setup/reconnect code around it.
B2: cmd/nomos/continue.go's processContinuations used to run every pending
continuation SEQUENTIALLY in a plain for loop, in the SAME goroutine as the
ticker — meaning (a) task B's continuation waited for task A's full (up to
10-minute) resumed turn to finish first, undercutting this session's earlier
concurrency work on exactly the path autonomous tasks depend on most, and
(b) an unrecovered panic anywhere in that call chain didn't just crash the
process (B1) — even WITH B1's recovery wrapped only at the top-level worker
spawn, the panic would still unwind the ENTIRE ticker-loop goroutine,
silently ending auto-continuation for every task until nomos restarted.
Fixed by spawning each pending item via safego.Go individually: real
parallelism, and a bad item can now only ever take down its own goroutine.
Added internal/safego/safego_test.go: TestGo_RecoversPanic is the concrete
proof — a deliberate panic inside Go() that would otherwise crash the whole
test binary; reaching the assertion after it IS the evidence recovery works.
Verified live against the rebuilt containers: full chat turn round-tripped
correctly (hostname lookup, 2 iterations, normal completion) — no regression
from threading safego.Go through the tool-dispatch/continuation paths.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Fix 1 of plans/2026-07-11-concurrent-task-execution.md — the safety-critical
one. The assent window (and destructive window) were keyed purely by agent
id ("assent_window.agent:<uuid>"). With one agent:nomos entity serving every
concurrent task, this meant approving Task A's plan opened a window that ANY
concurrently-running task's config-mutation/destructive actions could also
ride, auto-executing without their own approval.
- store.go / agent.go: assentWindowActive/openAssentWindow and
destructiveWindowActive/openDestructiveWindow/destructiveWindowKey all gain
a sessionID parameter; keys become
"assent_window.agent:<id>.session:<sessionID>" and
"destructive_window.agent:<id>.target:<slug>.session:<sessionID>". Missing
session id fails closed (no window) rather than falling back to the old
agent-wide key.
- continue.go: the auto-continuation worker's window check moved from once-
per-batch to once-per-pending-item, scoped to that item's own session —
it was previously checking ONE agent-wide window for a batch that can span
multiple tasks.
- agent.go tool-dispatch: injects `_session_id` into a COPY of the wire args
sent to the MCP server (never into the args used for the emitted/logged/
persisted tool call, and never part of any tool's declared InputSchema —
invisible to the model) so the gating checks on the OTHER side of the
process boundary know which task is asking.
- internal/mcp/server.go: assentWindowActive/destructiveWindowActive/
classifyAndGate gain the same sessionID parameter, read from
args["_session_id"] at the three call sites (request_execution's
apt_upgrade/pct_create branches, and the shared classifyAndGate used by
restart/pct_exec/systemctl/run).
Verified against the live stack with the exact scenario from the plan: opened
an assent window for session A only, then called `run` with an identical
config-mutation command for session A (window open) and session B (same
agent, no window). A auto-ran (execution status completed); B correctly
queued for approval (pending_approval) instead of bleeding through — proven
at both the MCP response text and the executions table.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Closes the gap that made the knowledge loop optional/implicit: every
non-trivial task now has an EXPLICIT first plan step (research) and last
plan step (write back), not just background behavior the model might skip.
New MCP tools (the agent had no way to do these before — only REST endpoints
existed, unexposed to it):
- update_entity_attributes(slug, attributes): shallow-merge new/changed facts
into an entity (an IP, a version, a discovered port) so a future task
doesn't have to rediscover them from scratch. No approval required — this
updates the knowledge graph, not live infra.
- create_relationship(source, target, type): record a discovered edge
(depends-on, hosts, provides, ...). Idempotent, FK-validated against the
ontology's relationship_types, no approval required.
SOUL.md: restructured the task loop so step 1 is explicitly "gather
knowledge, not just status" (get_entity_knowledge, search_knowledge,
get_relations, get_blast_radius, http_get) and the last step before
complete_task is explicitly "write back" (update_entity_attributes,
create_relationship, upsert_knowledge) — both called out as real plan
entries the operator should see in propose_plan, not silent side-work. This
is what prevents the graph drifting from reality and is the concrete
mechanism behind "tasks compound."
propose_plan's tool description reinforces the same first-step/last-step
convention at the call site.
Verified against the live stack: both tools registered and callable via MCP;
update_entity_attributes merged an attribute correctly; create_relationship
rejected an invalid type (FK violation, clear error) and succeeded with a
valid type+direction, confirmed idempotent (2 calls, 1 row).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Found live: a chat request to restart caddy (the reverse proxy for the
whole fleet) executed instantly over SSH with zero approval. Root
cause was in request_execution's legacy handler — restart, pct_exec,
and systemctl (outside enable/disable) executed immediately with a
hardcoded risk_class='reversible_low' that was never actually checked
against anything, bypassing the classifier entirely. Only the `run`
tool's commands were ever gated.
Extracted the run tool's classify -> execute-or-queue logic into a
shared classifyAndGate() and route restart/pct_exec/systemctl through
it too, so every mutating path — regardless of which tool the model
reaches for — gets the same read-only/config-mutation/destructive
classification and approval gate. systemctl restart is already covered
by an existing classifier test (config_mutation), so no new test
needed; the gap was that request_execution never called the
classifier at all.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
From the last (successful) TypeType deploy session, two gaps the operator hit:
1. Knowledge write-back — the missing half of the loop.
The agent could read the knowledge base (search_knowledge/get_entity_knowledge)
but had no way to WRITE it, so everything it learned (the Dragonfly memlock
rlimit gotcha, the NAT-hairpin DNS issue, etc.) lived only in an ephemeral
chat message and was lost — the system could never actually "get better."
This is the `upsert_knowledge` MCP tool the 2026-07-08 gaps plan called for.
- internal/mcp/server.go: upsert_knowledge(title, content, about?, tags?,
kind?) writes a document/investigation/runbook entity + knowledge_entities
row (search column is generated), upserts by slug so re-titling updates in
place, and optionally links it to the entity it's about so
get_entity_knowledge surfaces it there.
- SOUL.md: capture non-obvious findings/deploys/gotchas as part of finishing
work, not only when asked "what did we learn".
2. "I had to ask for status multiple times."
The clearest cause: a long working turn (64 tool calls) that exhausted the
iteration cap ended with a bare "max iterations reached without final
answer" — a dead end that forced the operator to ask what happened.
- cmd/nomos/agent.go: on exhaustion, make one final no-tools LLM call
(finalSummary) asking for a status report — what was accomplished, current
state, what remains — so the turn always ends with a real outcome.
- maxIterations 25 -> 40 (the decomposed per-step pct_create flow legitimately
needs more steps).
- SOUL.md: always end a turn with a clear outcome; never end silently or on a
bare tool call — the operator can't see the tools working and reads silence
as "nothing happened".
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The previous commit fixed a context-cancellation bug in the auto-approve path
and appeared to fix things, but re-testing end-to-end after deploy showed the
execution STILL never completed — just via a different symptom
("no pending execution found for approval" in the logs). Dug further and
found the real, deeper bug underneath: this whole mechanism has never
actually worked.
autoApprove() directly flipped BOTH approvals.status and executions.status to
'approved' via raw SQL, then called executeApprovedViaAPI to POST to the
decision endpoint. But DecideApproval's own logic specifically looks for the
execution still at status='pending_approval' to find and dispatch the real
SSH work (executeApprovedAction) — autoApprove's premature flip meant that
lookup always found zero rows. DecideApproval's UpdateApprovalStatus call
also silently no-ops the same way (sqlc :exec doesn't surface "0 rows
affected" as an error). Every assent-window auto-approved pct_create/
apt_upgrade has been sitting at 'approved' forever with the real work never
triggered — indistinguishable from "still running" until you check.
Fix: remove autoApprove() entirely. Call executeApprovedViaAPI directly
against the untouched pending_approval row from createApproval — identical
to the manual Approve-button path, just without the human click. DecideApproval
is now the single place that transitions status and dispatches, for both the
manual and auto-approved paths, closing the class of bug where two code paths
raced to do the same state transition.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Verified live testing the new atomic pct_create: an assent-window
auto-approved pct_create appeared to "run" (logged "auto-approved... running
now") but the execution stayed stuck at 'approved' forever. Root cause:
`go executeApprovedViaAPI(ctx, ...)` passed the MCP tool-call's own context —
which is cancelled the instant the triggering /chat request's HTTP response
completes, i.e. on every normal turn. The spawned goroutine's POST to the
approval-decision endpoint died with "context canceled" before it could even
start the real work, and nothing surfaced this to the operator or the agent —
the execution just sat at 'approved' with no error, indistinguishable from
"still running."
This is exactly the context-lifetime bug class httpapi's own approval
goroutine (executeApprovedAction) already avoided by using
context.Background() — it had just been missed in these two call sites
(apt_upgrade and pct_create auto-approve). Fixed both to use
context.Background(), matching the correct pattern already in place
elsewhere. Audited for other goroutines spawned with a request-scoped ctx —
none found; the sshExec internal goroutines are synchronous/waited-on via
select and correctly scoped to the call.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Closes the two remaining open points from the auto-continuation work.
1. Atomic pct_create (observability, the bigger of the two):
pct_create used to bundle create + apt install + post_install script into
one black-box multi-minute SSH call — the agent got back a single opaque
success/fail with no way to see (or fix) which step actually broke.
Removed the whole post-create provisioning block (and the now-dead
provisionScript/sanitizePkgs helpers + their tests). pct_create is now
create + start + register ONLY — fast, and its result is fed back to the
agent via auto-continuation almost immediately. The agent installs
packages and runs setup as its OWN sequence of `run` calls against the new
lxc:<hostname>, observing each command's real output and able to diagnose
and retry exactly the step that failed — the same recovery loop already
proven for the general case, now applied to installs too, instead of
requiring a separate black-box mechanism.
- services/post_install removed from the pct_create params struct and
from the MCP tool schema/SOUL.md docs.
- SOUL.md: explains the new flow, moves the Docker CLI gotcha and DNS
troubleshooting guidance to be steps the agent runs itself.
2. Scoped destructive window (targeted autonomy for recovery):
Verified live in the previous session that a destructive recovery (a
failed destroy needing stop-then-destroy on the same container) required
TWO separate typed confirmations for what was clearly one recovery
action. Added a narrow, TARGET-scoped 15-minute grant
(destructive_window.agent:<id>.target:<slug> in autonomy_settings,
shared key format across cmd/nomos and internal/mcp) that opens only
after an EXPLICIT typed confirmation (never loose assent) or an explicit
button-approval of a destructive step, and only ever covers further
destructive commands against that SAME target. A different target always
needs its own fresh confirmation — this narrows risk instead of loosening
it globally, unlike broadening the general assent window to cover
destructive actions would have.
- cmd/nomos/store.go: openDestructiveWindow/destructiveWindowActive/
executionTarget.
- cmd/nomos/agent.go: opens the window when a typed confirmation grants a
destructive chat-assent execution.
- internal/mcp/server.go: `run` tool checks the window before gating a
destructive command; auto-runs if active.
- internal/httpapi/phase3.go: DecideApproval opens the same window when a
destructive execution is approved via the button/API, for parity with
the chat-assent path.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
When the operator has approved a plan via chat assent (assent window
active), pct_create and apt_upgrade now auto-approve and execute
instead of queuing for a separate approval round. The auto-approve
path updates the approval+execution status in the DB, then calls the
HTTP API's decision endpoint to trigger executeApprovedAction — same
code path as a manual Approve button, consistent audit trail.
