Per the sequencing call (adversarial pass first, docs last), the docs now reflect the design the reviews confirmed. configuration.md: - Mutation ledger gains rows for SIGHUP reload, dashboard PATCH, dashboard reset, and the runtime overlay store, each mapping to the universal change protocol with what is and is not yet implemented. - New section 'Implemented runtime reload and config overlay (2026-08)': the Holder published-snapshot model, the reloadable-field criterion (a field is reloadable when EVERY consumer re-reads it), SIGHUP reload semantics, the overlay store (separate -config.sqlite, dotted keys, denylist, bootstrap degrade), and the PATCH/reset paths. safe-change-and-recovery.md: - New section 'Runtime config overlay: degrade paths and recovery' tabulating the three boot degrade paths and the reload degrade, all verified by the adversarial passes, plus the three documented recoveries (--no-config-overlay flag, rm the overlay file, dashboard reset). Status line notes the overlay subset is implemented while the full runtime-supervision protocol remains under design. No new documents, per the constraint.
41 KiB
Configuration
Status: Approved; existing-type assignment is ongoing
Date: 2026-07-28
Tracking issue: #73
Ownership
Configuration has two separate responsibilities:
internal/infrastructure/configurationloads files, environment variables, secret references, overlays, and compatibility representations.- Each domain and plugin owns the typed runtime settings required for its behaviour.
internal/app translates external input into those settings, validates
cross-capability combinations, and supplies each component only its own
settings.
The following are prohibited:
- passing complete application configuration into a domain;
- making a domain depend on TOML, YAML, environment-variable names, secret references, or a configuration library;
- defining a central runtime configuration type imported across the system;
- placing plugin options in an unrelated global configuration model.
Private input document
internal/infrastructure/configuration MAY use a complete decoded document
while discovering files, decoding sections, and applying overlays. It is an
input-boundary DTO only.
The document:
- remains private to the configuration infrastructure;
- is composed from separately defined feature sections;
- is immediately translated by application composition;
- is never returned as the application's runtime model;
- is never passed through runtime components.
External DTOs and domain runtime settings are separate types.
Plugins supply external configuration definitions through a narrow extension contract. The core loader MUST NOT gain a field or recognized filename for every plugin.
Invalid configuration input MUST NOT replace active valid settings or terminate unaffected capabilities. Configuration changes use the candidate validation, promotion, audit, and recovery requirements in safe-change-and-recovery.md.
Dissolution of internal/config
internal/config has no target location. It MUST be dissolved and MUST NOT
survive as a shared package, configuration domain, or renamed global model.
Its loading behaviour moves to internal/infrastructure/configuration. Existing
types and methods are reassigned by behaviour. Repository execution policy,
dispatch policy, retry policy, ecosystem defaults, tool policy, memory settings,
indexing settings, provider settings, and channel settings move to their
respective owners.
Repository settings
The existing config.Repo MUST NOT move intact into a repository domain or
shared settings package. A managed repository is an application association:
- work intake owns which repository references an identity monitors;
- task execution owns gates, preflight, test protection, ecosystem behaviour, and change limits;
- task lifecycle owns retry behaviour;
- scheduling owns concurrency policy;
- workspace management owns base branch and persistent-storage requirements.
Application composition correlates these settings by repository reference. No domain receives a universal repository profile.
Documentation
Generated configuration reference describes external sections and their owning capabilities. It MUST NOT establish a global Go configuration structure as the public contract.
Value classification rule
Every value that controls behaviour MUST have an explicit owner and MUST be classified as exactly one of:
- Invariant — changing it changes the meaning of the program, a wire protocol, or a domain concept. It may be a Go constant.
- Runtime setting — an operator, identity, domain, plugin, or deployment may legitimately choose a different value. It is typed configuration owned by the affected capability.
- Policy value — it expresses what is allowed, limited, routed, retried, protected, or gated. Its typed definition belongs to the owning domain and it is evaluated through the policy system.
- Derived value — it is computed from other settings and is not independently mutable.
- Runtime state — it changes as the application operates, such as the active model for a conversation. It is changed by a domain command and is not configuration merely because it can be changed.
An unclassified behavioural literal is a defect.
const MUST NOT be used merely because a value currently has one default.
Constants are reserved for actual invariants: enum discriminants, protocol
versions, fixed compatibility tokens, and values whose alteration would change
the program's semantics rather than configure an installation.
In Go, a constant is a compile-time value and does not represent mutable runtime storage. It generally has no address that can be observed or changed. Architecturally, a constant therefore asserts that the value cannot change for the lifetime of the application. Any value that may legitimately change while the application is running, between application instances, or between deployments is not a constant.
A runtime setting may have a default, but the default is not its only access
path. The owner exposes a typed settings structure, validation, and a
DefaultSettings constructor or equivalent. Private constants used to implement
defaults do not turn those defaults into invariants and MUST NOT be consumed
directly by runtime code.
Each runtime setting and policy value MUST declare:
- its owner;
- type, units, validation constraints, and safe bounds;
- default and derivation, if any;
- whether it is startup-only, live-reloadable, or restart-required;
- whether it may differ by installation, identity, repository, workflow, channel, or plugin instance;
- whether it is secret, sensitive, or safe to report;
- its current source, effective value, version, and last-known-good version;
- the command and identity required to change it.
Current configuration migration ledger
This ledger records the current meanings hidden inside internal/config. It is
the source inventory for dissolving that package; rows are split where the
existing structure combines different owners.
| Current values | Current location | Target owner and requirement |
|---|---|---|
WorkDir, DBPath |
root Config |
Workspace/storage infrastructure settings. DBPath is the prefix for independent task and conversation SQLite files. Defaults derive from the platform data directory. |
SkillsDir, PluginDir |
root Config |
Skill and plugin infrastructure settings, independently typed. |
PollInterval |
root Config |
Work-intake scheduling setting; not a daemon-wide constant. |
Label, BotUser, BotEmail |
root Config |
Identity and dispatch associations. BotEmail may be derived from the selected forge account but remains overrideable. |
MaxRetries |
root Config, Repo |
Retry policy owned by the operation being retried. The duplicate global/repository representation must disappear. |
DiffCapLines |
root Config |
Agent-change policy value. |
MaxSteps, MaxTokens, WallClock, GateMaxFailures |
Budgets |
Workflow-execution policy, with units and zero/unlimited semantics made explicit. |
Class, APIKeyEnv, BaseURL |
Provider |
Provider instance settings. Secret lookup becomes a SecretRef, not an environment variable name leaking into the provider domain. |
Mode, Command, Env |
Agent |
Agent-execution infrastructure settings. Arbitrary environment injection is explicitly trust-sensitive. |
Type, Host, Token / TokenEnv (older form) |
Forge |
Forge adapter settings. The older TokenEnv input is accepted only at the loader boundary. |
Trigger, AckReaction, label names |
Dispatch |
Work-intake and forge presentation policy. Label vocabulary is configurable per forge/identity association. |
Provider, ProviderConfig, SessionTTL |
MemoryConfig |
Memory capability settings. Provider-specific data remains owned and validated by that provider. |
Name, Transport, Command, Args, WorkDir, URL |
MCPServer |
One typed MCP server instance specification. Transport-specific variants must reject fields belonging to other variants. |
MaxResultChars, ParallelExecution |
ToolPolicy |
Tool-execution policy values, not generic daemon configuration. |
URL, TokenEnv |
NATSConfig |
NATS transport settings with a secret reference replacing the environment-name field. |
Enabled, Image, MaxConcurrency, MaxUptime, VolumeTTL, PullPolicy, Network |
ContainerConfig |
Container-execution infrastructure settings and resource policy. |
| model mappings, email, webhook, Telegram | ChatConfig |
Separate channel instance settings plus channel-routing associations; there is no chat-wide settings owner. |
AllowedUserIDs, TokenEnv, update commands |
Telegram config | Telegram adapter settings and access policy. Secret and policy fields are distinct. |
| listen/relay addresses | email config | Email adapter instance settings. |
| webhook target | Notify |
Notification adapter instance settings. |
| web listen address | Web |
Web transport settings. |
| index directory and database path | Indexing |
Index infrastructure settings, derived from storage settings unless explicitly overridden. |
| forge, dispatch, models, providers, agent, containers, memory, tools, indexing and repositories | IdentityConfig |
Existing composite duplicates unrelated settings. Replace it with associations from an IdentityID to independently owned configured instances. |
| owner, name, base, gate, protection, ecosystem, preflight, storage, retry and concurrency | Repo |
Split across work intake, workspace, workflow execution, policy, gate, and scheduling as described in Repository settings. |
TaskConfig, TaskForge |
worker snapshot | Replace with the minimum immutable, versioned WorkflowExecution input assembled by the application. It is not a second configuration model. |
Extra and arbitrary maps |
root/provider config | Allowed only at an extension boundary whose owning plugin/provider supplies schema and validation. Core configuration never consumes an untyped catch-all. |
Behavioural literals and incorrectly fixed values
The following current values affect useful behaviour but are not represented as owned typed settings or policy. They MUST be classified and migrated before the old configuration package is removed. The listed target is the architectural decision; exact setting names are implementation details.
| Capability | Values fixed in current code | Required classification |
|---|---|---|
| Chat agent | maximum steps 8 in cmd/archied |
Workflow/chat execution policy. |
| Context compression | enabled, threshold 0.5, context size 128000, protect first 3 and last 20, compression marker, token estimate of four characters per token |
Model/context runtime settings. The marker is presentation; the estimate is provider/model strategy, not a universal truth. |
| Tool guardrails | escalation thresholds 3, 5, 10, 2 |
Tool-use policy with named meanings and bounded defaults. |
| Tool output | turn budget 200000 characters and preview 500 characters |
Tool-execution resource and presentation settings. |
| Pairing/access | code TTL 1h, maximum pending 3, rate window 10m, failures 5, lockout 1h, maximum lockout 24h |
Access policy. Code length/alphabet and salt length require an explicit security-constraint decision; they are not automatically valid constants. |
| NATS daemon consumer | stream ARCHIE_TASKS, consumer archie-daemon, deduplication 2m, poll 2s, deliveries 3, acknowledgement wait 5m, inactive threshold 24h, subjects |
NATS topology settings plus delivery policy. Subject grammar or protocol headers may be invariants only if declared as versioned wire contracts. |
| NATS worker consumer | stream ARCHIE_TASKS, consumer archie-agent, deduplication 2m, poll 5s, acknowledgement wait 30m, deliveries 3, inactive threshold 24h, queue group and retry delay 1s |
Worker transport settings and delivery/retry policy. Shared stream defaults must have one owner rather than two copies. |
| Worker readiness | readiness timeout 20s, retry delay 250ms |
Worker lifecycle and retry policy. |
| Execution/RPC | agent execution 30m, task-run RPC 60s, worktree RPC 15m |
Per-operation timeout policy. |
| Indexing | build timeout 10m, metadata timeout 5s |
Index operation policy. |
| Plugin/tool cleanup | plugin rollback 10s, tool-provider cleanup 10s, MCP-provider cleanup 10s |
Lifecycle shutdown/rollback policy owned by each engine, with a shared policy vocabulary where appropriate. |
| Memory | sync buffer 100, prefetch 8s, shutdown 10s, single external-provider limit |
Memory resource/lifecycle settings. The provider-count restriction is a product rule only if deliberately retained. |
| Built-in memory store | file limit 100KiB, section-name length/character rules, query/match truncation limits, built-in provider/tool names |
Split into memory resource policy, input policy, presentation settings, and genuine registered identifiers. |
| Memory safety scanner | compiled prompt-injection and sensitive-data patterns with fixed warn/block consequences | Versioned memory-domain policy definitions. Patterns and consequences must be inspectable and replaceable without editing scanner code. |
| MCP stdio | initial backoff 500ms, maximum backoff 30s, unlimited default retries, send and shutdown 5s |
MCP transport retry, request, and lifecycle policy. Unlimited retries must be explicit rather than a zero-value accident. |
| MCP SSE/HTTP | reconnect 1s to 30s, POST/request timeout 30s |
MCP transport settings and retry policy. |
| Message framing | MCP frame 4MiB, worker protocol frame 8MiB, diagnostic 64KiB, media/base64 64MiB |
Resource/security policy unless a published wire protocol mandates the limit. Limits must be configurable only within safe bounds. |
| Webhooks and web UI | webhook bind 0.0.0.0:8644, webhook shutdown 5s, web read-header 5s, web shutdown 2s |
Transport instance and lifecycle settings. |
| Telegram | typing refresh 4s, draft interval 700ms, update-action TTL 10m |
Telegram adapter presentation and lifecycle settings. |
| Telegram limits | message length 4000 and callback/update prefixes |
Record the upstream platform constraint and version it as an adapter compatibility limit; prefixes may remain wire constants when changing them would invalidate outstanding callbacks. |
| Containers | stop and pull timeouts 10s and 30s |
Container lifecycle policy. |
| HTTP clients | feasibility notification timeout 10s; other paths use http.DefaultClient |
Adapter request policy. Every outbound client requires an explicit timeout profile. |
| Workflow routing | labels bug, feature, bootstrap; workflows tdd, feasibility, bootstrap; fallback implement |
Agent-system routing policy. Workflow identifiers are references, not compiled routing decisions. |
| Filesystem/deployment | .agents/skills, .archie/stages, .archie/gate.go, /data/worktree, /data/repo, cache names/paths and prepared sentinel |
Split into deployment settings, repository conventions, and genuine internal protocol markers. Paths an installation may relocate are settings. |
| Packaged release data | /usr/share/archie/CHANGELOG.archied.md and /usr/share/archie/CHANGELOG.archie.md |
Packaging/deployment settings or paths derived from an installation prefix. |
| Forge presentation | default label colour bfd4f2 and state-label colours |
Forge adapter presentation settings. |
| Fallback system prompt | compiled fallback prompt | Agent-system default content, versioned and replaceable through its owning settings; safety constraints remain policy/code. |
| Default paths and identities | config path, data paths, listen addresses, model roles and workflow names | Typed defaults belonging to their respective owners, never global constants consumed throughout the application. |
This inventory is intentionally broader than exported Go const declarations. A
literal hidden in a constructor, timeout, switch, retry loop, path join, prompt,
label map, or client setup restricts behaviour just as effectively as an
incorrectly used constant.
Constants that may remain
A value may remain constant only when its owner documents why runtime variation would be invalid. Typical valid cases are:
- domain enum values and state names;
- a versioned wire-protocol version or framing token;
- reserved internal headers and callback prefixes that peers must interpret identically;
- compile-time mathematical relationships;
- sentinel errors and internal type identifiers.
Names such as a stream, consumer, label, workflow, file, directory, network, model role, timeout, retry count, capacity, size limit, colour, prompt, command, or executable are not invariants merely because the current application has only one of them.
Architecture tests SHOULD reject exported behavioural constants outside approved
invariant packages and SHOULD detect direct consumption of private default
constants outside their owning DefaultSettings construction.
Current mutation and reload ledger
The application currently changes behavioural values through unrelated paths:
| Current change seam | Current behaviour | Required target |
|---|---|---|
Telegram /restart in cmd/archied |
Re-reads the global document and directly replaces Telegram token and allowed-user IDs; failure retains old values. | Submit a Telegram-settings candidate through the universal change protocol. Validate the complete candidate, promote atomically, audit the acting identity, and retain last-known-good. |
| Active model selection | chatModelManager.SetActiveModel changes a model reference. |
A conversation/session command changing runtime state. Validate against configured model instances and audit it; do not rewrite configuration. |
| Active provider selection | chatModelManager.SetActiveProvider changes provider selection. |
A conversation/session command changing runtime state, with the same distinction from provider-instance configuration. |
| Persona selection | PersonaRegistry.SetActive mutates an in-memory per-session choice. |
Agent-system session state. Persistence and audit requirements are owned by that domain, not the configuration loader. |
Memory provider SaveConfig |
Accepts an untyped map, delegates validation/persistence to an optional provider interface, then providers reload backing files. | A typed provider-owned configuration command using candidate validation, versioning, atomic promotion, health observation, audit, and rollback. |
| File and environment overlays | Applied while loading internal/config; defaults are filled by mutation in finalize. |
Configuration-infrastructure input processing that produces separately typed candidates. Defaults and derivations remain visible in provenance. |
| SIGHUP config reload (2026-08) | Re-resolves the file config (+ the runtime overlay), validates, then atomically republishes through config.Holder. A failed reload keeps the running config and records last_error/last_error_at in /api/config. |
Candidate promotion with last-known-good retention: the running snapshot is untouched on failure. A bounded observation period and actor audit are not yet implemented. |
Dashboard PATCH /api/config (2026-08) |
Applies dotted-key updates to a deep copy of the published config, validates the materialised result, persists to the runtime overlay, then republishes. Denylisted keys (db_path, work_dir) return 400; a rejected candidate never reaches Set. |
The universal change protocol's validate-stage-promote path with the overlay store as persistence. Actor is recorded as dashboard; no per-field health observation or versioned audit trail yet. |
Dashboard reset POST /api/config/reset (2026-08) |
Deletes one overlay row after resolving and validating the target state, then republishes file + remaining overlay, so a dashboard-created override can be removed without editing SQL. | Remediation/rollback: the last-known-good (file) value is restored for that key. |
| Runtime config overlay store (2026-08) | Dashboard-edited overrides in cfg.DBPath + "-config.sqlite" (own user_version, own migrator), layered over the file config at boot and on every reload. A broken overlay degrades to file config alone, surfaced via ReloadStatus.OverlayUnavailable. |
External configuration storage for runtime-tunable values, separate from the task/conversation stores so each file owns its migration. The --no-config-overlay / ARCHIE_SKIP_CONFIG_OVERLAY=1 recovery hatch skips it entirely. |
Any further mutation seam discovered during migration MUST be added here before it is changed. Direct field assignment on a live settings object is prohibited.
Universal change protocol
Changing configuration is an application command performed by an identity. It MUST:
- resolve the target owner and current version;
- parse into that owner's typed settings;
- validate syntax, semantic constraints, policy, references, and safe bounds;
- stage a candidate without altering the active value;
- run owner-defined readiness and health checks;
- atomically promote the candidate and record actor, source, reason, diff, prior version, and new version;
- observe the affected capability for a bounded period;
- retain the new version on success, or automatically restore last-known-good on failure;
- report the rejected or rolled-back candidate without terminating unaffected capabilities.
Startup loading uses the same candidate rules. A bad user value therefore does not partially mutate the application or make a previously healthy capability unavailable.
Settings that cannot safely reload in-process are marked restart-required. The change may still be validated and versioned immediately, but promotion occurs through a controlled restart with the same health and rollback semantics.
Implemented runtime reload and config overlay (2026-08)
A first implementation of the universal change protocol now exists for the runtime-tunable surface: SIGHUP reload, the dashboard PATCH, and the runtime config overlay. The mechanics below are the current implementation; the aspirational protocol (bounded observation, versioned audit trail, per-actor audit) remains future work.
Published snapshot: config.Holder
The running configuration is a published snapshot held in config.Holder
(internal/config/holder.go). Get returns a value under RLock; Set replaces the
whole snapshot. A published Config is immutable: reload constructs a fresh one
and calls Set. config.Clone deep-copies every reference-type field so a failed
decode cannot mutate the published snapshot's shared maps. Holder.Get panics
on a nil receiver — a forgotten NewHolder must fail loudly, not boot a daemon
that silently runs on an empty config.
The reloadable-field criterion
The diff-and-warn on reload asks whether a changed field takes effect without a
restart. The criterion is: a field is reloadable when every consumer re-reads
it. That question read as "what does the daemon re-read" while the daemon owned
the only Holder; it stopped being sufficient when webui began sharing that
Holder, because webui handlers re-read on every request. The allowlist lives in
cmd/archied/reload.go and is deliberately a deny-by-default list: a field added
to Config later defaults to requires-restart, forcing whoever adds it to decide
deliberately. A reload that changes any non-allowlisted field logs
"requires a restart" with the field names; a fully-applied reload logs Info.
SIGHUP reload
SIGHUP re-resolves the file config (plus the runtime overlay), validates, and
atomically republishes through the shared Holder. A failed reload — bad file,
bad overlay — keeps the running config and records last_error/last_error_at
in /api/config so the stale state is visible where the operator is looking.
The three poll tickers re-read PollInterval per iteration and reset on change,
so a reloaded interval genuinely takes effect.
The runtime config overlay
The dashboard's edits persist to a dedicated SQLite file
(cfg.DBPath + "-config.sqlite", its own user_version and migrator — no
contention with the task or conversation stores). The config_overlay table
stores dotted-path keys with JSON-encoded values; the daemon layers them over
the file config at boot and on every reload, with the same field-level
precedence as a file overlay.
- Denylist:
db_pathandwork_dircannot be set (the daemon must read them before it can open the overlay store; changing them could break the next boot). Enforced at write time with a 400 and in the store. - Bootstrap degrade: a broken overlay (unopenable store, unreadable value,
values failing validation) degrades to file config alone — it must not brick
the daemon. The reason is surfaced via
ReloadStatus.OverlayUnavailablein/api/configand rendered as a banner by the dashboard. - Recovery hatch:
--no-config-overlay/ARCHIE_SKIP_CONFIG_OVERLAY=1skips the overlay entirely; removing the-config.sqlitefile is the documented manual recovery. See safe-change-and-recovery.md. - Shadowed file edits are visible:
/api/configreports which dotted keys the overlay currently sets (overridden), so the dashboard marks those rows and offers a per-row reset.
Dashboard PATCH and reset
PATCH /api/config receives {"updates": {"dotted.key": value}}. It applies the
updates to a deep copy of the published config, validates the materialised
result, persists to the overlay, then republishes — it never calls Set directly
from the handler, since the dashboard shares the daemon's Holder. Dotted keys
are nested before decode (yaml cannot decode a dotted key as a struct field).
Error classification: 400 for a denylisted key or failed validation, 503 when
the overlay is disabled, 500 for a persistence failure. POST /api/config/reset
deletes one overlay row after resolving and validating the target state,
restoring the file value — the answer to "my file edit isn't taking effect
because an override shadows it".
Completion criteria
The configuration rearchitecture is complete only when:
internal/configno longer exists;- every behavioural value has one classification and owner;
- every current field and hard-coded value in the ledgers has a recorded disposition;
- domains and plugins receive only their typed settings;
- all changes use semantic commands and the universal change protocol;
- active, candidate, rejected, rolled-back, and last-known-good versions are queryable with secret-safe provenance;
- generated documentation lists every external key, default, constraint, mutability class, owner, and deprecation;
- architecture tests prevent a new central runtime configuration model, direct live-setting mutation, and incorrectly fixed behavioural values.
Startup policy: missing credential degrades, invalid config is fatal
Decided 2026-07-30. Recovered 2026-08-09 from the pre-migration issue tracker.
A missing credential disables a capability. Only an invalid config stops the daemon. "Forge is a feature, not a requirement."
Before 518a1a0, an unresolvable forge token did return 1 in
cmd/archied/main.go, so a chat-only user who never configured GitHub got no
chat, no gateway, nothing — and under a systemd unit with Restart=on-failure
and no start limit it crash-looped every five seconds with the error scrolling
past unread. Forge construction now lives in resolveForge, shared by the
primary forge and every configured identity, and degrades to forge.NewNoop
with a warning. The identity path had the same defect and was worse: one
identity's missing credential killed all the others.
Malformed configuration still fails fast in configuration.Validate, well
before this point. That distinction is the whole design.
Apply the same reasoning to any future capability credential, LLM keys
especially: start, report the capability as unavailable, let the operator see it.
Guarded by TestResolveForgeDegradesInsteadOfFailing.
Known consequence, tracked separately: worktree.Manager still receives the
empty token, so pushes fail late rather than at startup.
Config generation belongs to archied setup, not install.sh
Decided 2026-07-30 (#368). Recovered 2026-08-09 from the pre-migration issue tracker.
install.sh must stop generating config.toml; archied setup owns it. The
installer hand-rolling TOML has already drifted from the schema once: the
generated-config branch hardcoded key = ARCHIE_GITHUB_TOKEN and
host = ${GITEA_URL} regardless of the answers given, so choosing Gitea produced
a config naming a token the installer never wrote (archied then exited 1), and
choosing GitHub pointed host at gitea.example.com. 61a5b35 patched this
with a forge_block shell helper shared by both config paths — that helper is a
workaround, deleted by
#72. The durable fix is
that the code writing the config is the code that reads it.
Why Go and not Python: the installer must run where nothing is installed yet, and
python3 with tomllib means 3.11+, which is not safe on LTS distros. archied
is already being built two steps earlier.
Key constraint for the port: BurntSushi/toml does not survive a
decode/encode round trip with comments intact, and config.example.toml carries
documentation users are meant to hand-edit. Decide the write strategy
(#412) before writing
any prompting code. Do not port a separate config.yaml/auth.yaml secret
split — Archie already has secret.SecretRef{engine,key}, so setup writes the
reference to TOML and the value to the secret engine.