Reproducibility

A research-quality release should be reproducible from a clean checkout. The public repository should therefore publish a clean snapshot, not a private development history containing internal notes.

Environment

• Rust toolchain: 1.88.0
• Docs package manager: pnpm
• Desktop package manager: pnpm
• Default database-backed tests use PostgreSQL only when TEST_DATABASE_URL or ASTEREL_POSTGRES_URL is configured.

Minimal release evidence

For the strict release gate, use the checked-in release script so quality, supply-chain, baseline, and replay checks stay in one source of truth:

./scripts/release/human_like_release_gate.sh

When the full strict gate is too expensive for a local documentation pass, record that limitation and run the closest targeted subset:

cargo fmt -- --check
cargo clippy -- -D warnings
cargo check-all
cargo test
docker compose config
pnpm --dir docs build
cargo metadata --no-deps --format-version 1

Research evidence snapshot

For a paper-facing artifact, record:

• public commit hash;
• Rust, Node, pnpm, and OS versions;
• command list and pass/fail result;
• fixture file hashes;
• model/provider versions for any non-deterministic evaluation;
• random seeds and sampling parameters;
• redaction policy used for any logs or transcripts;
• known ignored tests and why they were excluded.
• whether the strict release gate or a documented targeted subset was run.

Clean snapshot requirement

Because the private development history may include internal review notes, local agent assets, and operational handoffs, public initialization should use a clean snapshot of the public tracked set. Historical private blobs are not research evidence and should not be preserved in the public repository history.

Use the snapshot helper to avoid copying ignored caches or local-only files:

scripts/release/create_public_snapshot.sh /tmp/asterel-public-snapshot --dry-run
scripts/release/create_public_snapshot.sh /tmp/asterel-public-snapshot