Sibyl 1.0 RC Release Packet

• Status: candidate prepared; external release receipts pending
• Version: 1.0.0-rc.1
• Branch: feature/sibyl-1-0-rc-plan
• Release floor: v0.10.0
• Rollback target: v0.10.0

This packet records the repo-side evidence for the v1.0 RC candidate. The candidate SHA is the final committed SHA on this branch at release time. The release workflow records that SHA in the rc-gate-receipt-* artifact before it can tag or dispatch publishing.

Do not tag or publish until Bliss gives explicit release go-ahead.

Decision Gate

Ship now: no.

Smallest blocker: same-SHA GitHub CI and Nightly Regression must run on the final candidate SHA, then the release workflow must be dispatched with that Nightly Regression run ID.

Residual release risk: Docker image startup, public PyPI pages, GitHub release body, container manifests, and clean installs from published artifacts are post-publish checks.

Claim Matrix

Claim	Receipt	Status
Active docs agree	moon run docs:lint and moon run docs:build	Local pass
Task graph is current	sibyl epic show epic_19e1dea67ebf; sibyl task list --status doing; no RC todo or blocked tasks	Local pass
Source ingest is current	moon run adapter-ingest-gate; moon run large-corpus-rehearsal	Local pass
Synthesis is source-grounded	moon run synthesis-gate	Local pass
Automatic memory is policy-safe	moon run autonomy-gate; moon run memory-trust-gate; moon run trust-control-gate	Local pass
Sessions are boring	moon run auth-session-gate	Local pass
Reflection quality is current	moon run reflection-quality-gate	Local pass
Context and workspace trust are current	moon run context-quality-gate; moon run workspace-trust-gate	Local pass
Overview performance is current	moon run overview-perf-gate	Local pass
Surreal-only runtime holds	moon run inventory-check; moon run inventory-typecheck; moon run inventory-test; supported grep audit	Local pass
Redis is optional locally	moon run api:test -- tests/test_coordination_local.py -v; moon run api:memory-trust-jobs-test	Local pass
Backup/restore is release-gated	moon run backup-restore-gate	Local pass
Benchmark ledger is claim-safe	moon run bench-gate; Nightly Regression compare artifacts	Nightly pending
Package artifacts build	moon run python-package-build produced sibyl-core, sibyl-dev, and sibyld wheels and sdists	Local pass
Package installs work	Clean isolated uv tool install for sibyl-dev and sibyld; both entrypoints report 1.0.0rc1	Local pass
Helm surface works	helm lint charts/sibyl; helm template sibyl charts/sibyl has no graphiti, falkor, or postgres hits	Local pass
Release cut is gated	.github/workflows/release.yml runs moon run :check before tag/publish and validates same-SHA nightly	Workflow guard
Publish dispatch is gated	.github/workflows/publish.yml runs moon run :check before Python and Docker artifacts	Workflow guard
Rollback is ready	Roll back to v0.10.0; do not move tags; verify published artifacts before announcing RC	Operator action

Receipt Highlights

• moon run :check -> 47 completed.
• moon run release-workflow-test -> 6 passed.
• moon run python-package-build -> built sibyl_core-1.0.0rc1, sibyl_dev-1.0.0rc1, and sibyld-1.0.0rc1 artifacts.
• Clean isolated install -> sibyl 1.0.0rc1; sibyld 1.0.0rc1.
• moon run docs:build -> build complete, with existing Rollup chunk and PURE annotation warnings.
• helm lint charts/sibyl -> 1 chart linted, 0 failed.
• Helm render -> 350 lines; no graphiti, falkor, or postgres references.

Release Dispatch Requirements

1. Push feature/sibyl-1-0-rc-plan.
2. Run GitHub CI on the final candidate SHA.
3. Run Nightly Regression on the final candidate SHA and keep its baseline parity, live graph, and benchmark artifacts.
4. Dispatch Release with version=1.0.0-rc.1, dry_run=false, and the Nightly Regression nightly_run_id.
5. Verify the GitHub release, PyPI package pages, Docker manifests, docs install page, and clean installs from published artifacts before announcing the RC.