Omnigraph v0.6.0
Three pieces of work land in this release:
Three pieces of work land in this release:
- The graph terminology rename (renamed
Repo→Graphacross the Cedar resource model, policy API, and query-lint schema source). - Multi-graph server mode — one
omnigraph-serverprocess can now serve 1–10 graphs concurrently behind cluster routes (/graphs/{graph_id}/...), with per-graph and server-level Cedar policy, read-onlyGET /graphsenumeration, and CLI parity (omnigraph graphs list). - Inline + canonical-named queries and mutations. New
POST /queryandPOST /mutateendpoints pair with the CLI's new-e/--query-stringflag for ad-hoc execution without a temp file.POST /readandPOST /changecontinue serving indefinitely as deprecated aliases that carry RFC 9745Deprecation: trueand RFC 8288Link: </successor>; rel="successor-version"response headers, plusdeprecated: trueinopenapi.json. Same canonicalization on the CLI:omnigraph query,omnigraph mutate, and top-levelomnigraph lint/omnigraph checkreplaceomnigraph read,omnigraph change, and the nestedomnigraph query lint/omnigraph query check. Every deprecated spelling remains avisible_aliasthat warns to stderr once per invocation.
Runtime add/remove (POST /graphs, DELETE /graphs/{id}, omnigraph graphs create) is not in v0.6.0. Operators add or remove graphs by editing omnigraph.yaml and restarting. The first cut of POST /graphs shipped behind an atomic-YAML-rewrite design that we pulled before release once its concurrency guarantees were challenged (flock-on-renamed-inode race, duplicate-check outside the critical section, and an init-cleanup path that could destroy an existing graph's schema on re-init). The correct fix is a Lance-style cluster catalog (reserve → init → publish with recovery sidecars); that work is deferred.
Breaking Changes
Graph terminology rename
- Renamed the Cedar resource entity from
Omnigraph::RepotoOmnigraph::Graph. - Renamed policy API terminology from
repo_idtograph_idonPolicyCompiler::compile(and on the newPolicyEngine::load_graph/PolicyEngine::load_serverloaders described below). - Renamed query-lint schema source JSON from
"repo"to"graph"forschema_source.kind.
Multi-graph server mode
- Multi-graph deployments lose flat routes. Single-graph invocation (
omnigraph-server <URI>) is unchanged — same flat/snapshot,/read,/branches, etc. Multi-graph deployments serve those routes under/graphs/{graph_id}/...; bare flat paths return 404 in multi mode. ServerConfigshape change (programmatic embedders only):ServerConfig { uri, policy_file }is replaced byServerConfig { mode: ServerConfigMode }, whereServerConfigMode = Single { uri, policy_file } | Multi { graphs, config_path, server_policy_file }. Callers that useload_server_settingsare unaffected; callers that constructServerConfigdirectly need to wrap their fields inServerConfigMode::Single.AppState's routing surface isAppState::routing() -> &GraphRouting, whereGraphRouting = Single { handle } | Multi { registry, config_path }. The previousAppState::uri(),AppState::mode(),AppState::registry()accessors and theServerModeenum are gone — embedders readstate.routing()and match on the arm they need. Per-graph URIs live onhandle.uri.AppState::new_multiis the new multi-graph constructor. Single-modenew_*/open_*constructors are unchanged.AuthenticatedActor(Arc<str>)→ResolvedActor { actor_id, tenant_id, scopes, source }(programmatic embedders only). The struct shape changes, but the HTTP contract — bearer auth and the bearer-derived-actor-identity guarantee — is unchanged. Cluster-mode call sites construct withtenant_id: None,scopes: vec![Scope::Full],source: AuthSource::Static. The new fields are forward-compat seams for future multi-tenant and OAuth deployments; they're inert in this release.PolicyEngine::load(path, graph_id)removed in favor of two kind-typed loaders:PolicyEngine::load_graph(path, graph_id)for per-graph policies andPolicyEngine::load_server(path)for server-level policies. Each loader rejects rules whose actionresource_kind()doesn't match the engine kind — operators who put agraph_listrule in a per-graph file (or areadrule in a server file) now get a load-time error instead of a silently-never-matching rule.PolicyRequest::actor_idfield removed. Actor identity is now a separate parameter onPolicyEngine::authorize(actor_id, &request). The type system enforces the server-authoritative-actor invariant: actor identity is always sourced from the bearer-token match resolved at the auth boundary; handlers cannot smuggle identity through the request body.Omnigraph::initis strict by default. Initialization at a URI that already holds schema files now errors withOmniError::AlreadyInitializedinstead of silently overwriting. Operators who actually want to overwrite useInitOptions { force: true }(CLI:omnigraph init --force). Closes the destructive-cleanup footgun where a failed re-init would delete an existing graph's schema files.- Top-level
policy.fileis rejected in multi-graph server mode. It remains valid for single-graph / CLI-local policy. Multi-graph deployments must move graph rules tographs.<graph_id>.policy.fileand server-scopedgraph_listrules toserver.policy.file. - Open server startup requires explicit opt-in. A server with no bearer tokens and no policy now refuses to start unless passed
--unauthenticatedorOMNIGRAPH_UNAUTHENTICATED=1. - Policy requires bearer tokens. Configuring any policy file without bearer tokens now refuses startup; otherwise every protected request would 401 before Cedar could evaluate it.
- Tokens without policy default-deny non-read actions. Existing authenticated deployments that relied on writes or admin routes without Cedar policy must add policy rules for those actions.
GET /graphsrequiresserver.policy.filein every runtime state. Even--unauthenticatedmode keeps server topology closed until the operator explicitly authorizesgraph_list.
Query / mutation rename
ChangeRequestfield rename:query_source→query,query_name→name. Both legacy names continue to deserialize via#[serde(alias = "...")], so existing clients sending the old JSON keys keep working. CLI remote calls against/changestill emit the legacy keys verbatim through thelegacy_change_request_bodyhelper so a newer CLI talking to an older server keeps working byte-for-byte.- CLI
omnigraph query lint/omnigraph query checkare now top-level — canonical name isomnigraph lint. The three deprecated invocations (omnigraph query lint,omnigraph query check, and bareomnigraph check) remain as argv-level shims that rewrite toomnigraph lintand print a one-line stderr deprecation warning.checkis deliberately not a clapvisible_aliasonlint— two equivalent canonical names would split agent emissions between them depending on training-data drift, so the deprecation pattern (rewrite + warn) gives one unambiguous canonical name inomnigraph --help.
New
- Multi-graph mode. Invoke with
omnigraph-server --config omnigraph.yamlwhere the YAML has a non-emptygraphs:map and no single-mode selector (noserver.graph, no CLI<URI>or--target). At startup the server opens every configured graph in parallel (bounded concurrency, fail-fast). GET /graphs. Lists every registered graph, sorted alphabetically bygraph_id. Auth-required when bearer tokens are configured; Cedar-gated byPolicyAction::GraphListagainstOmnigraph::Server::"root". Returns 405 in single mode. Server-scoped actions require an explicitserver.policy.filein every runtime state — the management surface is closed by default even in--unauthenticatedmode so that server topology is never exposed without operator opt-in.- CLI
omnigraph graphs list. Mirrors the HTTP surface. Rejects local URI targets with a clear message — for remote multi-graph servers only. - CLI
omnigraph init --force. Bypasses the strict-init preflight when an operator deliberately wants to recover from orphan schema files. Does NOT purge existing Lance datasets; recursive deletion needsStorageAdapter::delete_prefix(deferred — see below). - Per-graph Cedar policy. Each entry in the
graphs:map can carry apolicy.filepath, loaded at startup viaPolicyEngine::load_graph. Cedar'sOmnigraph::Graph::"<graph_id>"resource is per-graph; the newOmnigraph::Server::"root"resource governs server-level actions. - Server-level Cedar policy.
server.policy.filein the config governs thegraph_listaction onOmnigraph::Server::"root". Required to exposeGET /graphsin every runtime state — without a server policy the default-deny posture rejectsgraph_list, including in--unauthenticatedmode. - Cedar action vocabulary:
graph_list(server-scoped). Runtimegraph_create/graph_deleteare reserved but not shipped — see "Deferred." - Canonical graph URI identity. Server startup normalizes graph root URIs before registry insertion and response output, so aliases such as
/tmp/g,/tmp/g/, andfile:///tmp/gcannot register as distinct graphs that actually share one Lance root. POST /queryandPOST /mutate. Canonical inline endpoints./queryrejects mutations with a typed 400 (the D2 rule lives at the URL — read-only contract enforced before execution); body uses the clean{ query, name, params, branch, snapshot }shape./mutateaccepts the same shape for mutations. Both available in single mode and per-graph multi mode (/graphs/{id}/query,/graphs/{id}/mutate). Internal call sites share two helpers (run_query,run_mutate) that take decoupled args, not request bodies — the seam MR-969's future stored-query handler plugs into.- CLI
omnigraph query/omnigraph mutateas top-level canonical subcommands. Pairs with new top-levelomnigraph lint(aliascheck) so query validation no longer sits underomnigraph query. - CLI
-e, --query-string <GQ>on bothomnigraph queryandomnigraph mutate. 3-way mutex with--query <path>and--alias <name>— exactly one is required. Empty string rejected. Suits ad-hoc exploration, REPL workflows, and agent tool-use without temp files. - Three-channel deprecation signal on
/readand/change: OpenAPIdeprecated: trueon the operation (every codegen flags the generated SDK method), RFC 9745Deprecation: trueresponse header, and RFC 8288Link: </query>; rel="successor-version"(or</mutate>) response header. Auto-discoverable; no SDK breakage. omnigraph.yamlaliases.<name>.commandnow acceptsqueryandmutateas canonical values alongside the legacyreadandchange. The internalAliasCommandenum retains the legacy variant names so serialized configs stay byte-stable.
Configuration
omnigraph.yaml schema additions (all optional, single-mode unaffected):
server:
bind: 0.0.0.0:8080
policy:
file: ./server-policy.yaml # server-level Cedar (graph_list)
graphs:
alpha:
uri: s3://tenant-bucket/alpha
policy:
file: ./policies/alpha.yaml # per-graph Cedar
beta:
uri: s3://tenant-bucket/beta
# no per-graph policy → engine-layer enforcement is a no-opDeferred
POST /graphsruntime graph creation and CLIomnigraph graphs create. Pulled before release after the YAML-rewrite design's correctness story didn't survive review. A future release will add a managed cluster catalog (Lance-backed reserve → init → publish with recovery sidecars) and re-expose runtime creation on top of it. Until then, operators add graphs by editingomnigraph.yamland restarting.DELETE /graphs/{id}. Never shipped in v0.6.0; deferred with the same cluster-catalog work.StorageAdapter::delete_prefix. The substrate primitive a managed catalog would need. Will land alongside runtime mutation.omnigraph init --forcepurging Lance state. Today--forceonly bypasses the schema-file preflight; recursive deletion of existing Lance datasets needsdelete_prefix.X-Actor-Idservice delegation forwarding. Needs durable both-actor audit on_graph_commits.lance— out of scope.- Hot policy reload. Restart is cheap at N≤10 graphs.
User Impact
- No on-disk migration is required. Existing
.omnigraphs from v0.5.0 (and earlier) open cleanly under v0.6.0 — Lance datasets,__manifest,_schema.pg,_schema.ir.json,__schema_state.json,_graph_commits.lance,_graph_commit_recoveries.lanceall use unchanged formats. No conversion step. - Existing single-graph storage upgrades without migration. Server deployments may need auth/policy config changes: explicitly pass
--unauthenticatedfor local open mode, configure tokens when using policy, and add Cedar policy for non-read authenticated actions. - Multi-graph adoption is opt-in. Add a
graphs:map toomnigraph.yaml(and removeserver.graph) to switch a deployment to multi mode. - Cluster routes are breaking for client SDKs targeting multi mode. Generated clients from previous v0.5.0 OpenAPI specs will hit 404 on flat paths against a multi-mode server. Regenerate against the v0.6.0
openapi.json. - Supported YAML policy authoring is unchanged. The Cedar
Omnigraph::GraphandOmnigraph::Serverentities are internally generated bycompile_policy_source— operator YAML only references actions and groups. - Operators with unsupported raw Cedar policy files should update
Omnigraph::Reporesource references toOmnigraph::Graph. - Endpoint and CLI rename is cosmetic on the client side. Existing callers on
/read,/change,omnigraph read,omnigraph change, andomnigraph query lintkeep working — they pick up theDeprecation+Linkheaders (or stderr deprecation warning on the CLI) so SDKs and proxies can surface the successor name automatically. New integrations should target the canonical names. ChangeRequest field names migrate at the caller's pace — bothquery_source/query_nameandquery/nameaccepted indefinitely.
Migration: single → multi
# Before (v0.5.0 single-mode invocation)
server:
graph: my-graph
graphs:
my-graph:
uri: /var/lib/omnigraph/my-graph
policy:
file: ./policy.yaml# After (v0.6.0 multi-mode — drop `server.graph` and the top-level `policy`)
server:
policy:
file: ./server-policy.yaml # NEW: governs GET /graphs
graphs:
my-graph:
uri: /var/lib/omnigraph/my-graph
policy:
file: ./policy.yaml # MOVED: was top-levelSame omnigraph.yaml file; restart the server. Clients targeting the old flat routes (/snapshot, /read, …) must update to /graphs/my-graph/snapshot, etc.
To add a new graph after rollout: stop the server, append a new graphs.<id> entry, restart.
Documentation
- Public docs, CLI help, examples, server docs, and test helpers now consistently use "graph" for the OmniGraph data artifact.
- GitHub/source repository terminology remains spelled out as "repository" where needed.
- New:
docs/user/cli.mddocumentsomnigraph graphs list;docs/user/server.mddocuments the multi-graph mode and the cluster route convention;docs/user/policy.mddocuments the per-graph vs server-scoped action distinction. - New:
docs/user/server.mddocumentsPOST /query/POST /mutateand the three-channel deprecation signal on/read//change.docs/user/cli.mddocuments the-e/--query-stringflag with examples.docs/user/cli-reference.mdshows the canonical CLI verbs (query,mutate,lint,check) with legacy spellings as visible aliases. - New:
docs/dev/rfc-001-queries-envelope-mcp.mdis the cross-cutting design doc for the inline / stored query work that started landing in this release. It sequences the v0.6.x patch series (request/response envelope hardening) and the v0.7.0 stored-query + MCP work.
Test coverage
GraphIdnewtype validation, registry race tests, init failpoints (still reachable fromomnigraph initCLI).- Mode-inference four-rule matrix, parallel multi-graph startup, cluster routing.
- Cedar
Serverresource refactor, backwards-compat for graph-only policies, kind-alignment rejection (server actions in graph files / vice versa). GET /graphsenumeration, 405-in-single-mode, 403-in-Open-mode-without-server-policy, Cedar admin/viewer authorization.- Cluster routes with inner path params (
/branches/{branch},/commits/{commit_id}) deserialize correctly under axum 0.8 nested routing. - Policy-requires-tokens startup invariant enforced uniformly across single and multi mode.
- The bearer-auth-derived-actor-identity regression test (client-supplied identity headers are ignored; the server-resolved actor is the only identity Cedar sees) stays green across the entire refactor.