菜单
Use when validating a Shasta proposal locally from a Taiko L2 block height; discovers the proposal tuple, runs preflight, and verifies the generated GuestInput with the native guest launcher.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | shasta-proposal-regression |
| description | Use when validating a Shasta proposal locally from a Taiko L2 block height; discovers the proposal tuple, runs preflight, and verifies the generated GuestInput with the native guest launcher. |
Required:
network: L2 chain spec key, usually taiko_hoodi.l2_block: any L2 block height inside the proposal to validate.Optional:
l1_network: defaults to hoodi for taiko_hoodi.work_dir: defaults to /tmp/raiko2-shasta-regression-<l2_block>.NETWORK=taiko_hoodi
L1_NETWORK=hoodi
L2_BLOCK=7225500
WORK_DIR=/tmp/raiko2-shasta-regression-${L2_BLOCK}
mkdir -p "$WORK_DIR"
python scripts/regression/stress_shasta_proposal.py \
--network "$NETWORK" \
--l1-network "$L1_NETWORK" \
--l2-block-range "${L2_BLOCK},$((L2_BLOCK + 1))" \
--discover-only \
--proposal-out "$WORK_DIR/proposal.discovery.json"
proposal.discovery.json and use the first item in .proposals[].eval "$(
python - "$WORK_DIR/proposal.discovery.json" <<'PY'
import json
import sys
proposal = json.load(open(sys.argv[1]))["proposals"][0]
print(f'PROPOSAL_ID={proposal["proposal_id"]}')
print(f'L1_INCLUSION_BLOCK_NUMBER={proposal["l1_inclusion_block_number"]}')
print(f'LAST_ANCHOR_BLOCK_NUMBER={proposal["last_anchor_block_number"]}')
print(f'L2_START={proposal["l2_start"]}')
print(f'L2_END={proposal["l2_end"]}')
PY
)"
cargo run -r -p preflight -- \
--network "$NETWORK" \
--l1-network "$L1_NETWORK" \
--proposal-id "$PROPOSAL_ID" \
--l1-inclusion-block-number "$L1_INCLUSION_BLOCK_NUMBER" \
--last-anchor-block-number "$LAST_ANCHOR_BLOCK_NUMBER" \
--l2-start "$L2_START" \
--l2-end "$L2_END" \
--proof-type native \
--validate \
--output "$WORK_DIR/guest-input.json"
cargo run -r -p guest-launcher -- \
--proof-type native \
--mode prove \
--input "$WORK_DIR/guest-input.json" \
--output "$WORK_DIR/native-proof.json"
raiko2 server for this workflow.native; do not switch this workflow to SP1 or SGX.--l1-rpc, --l2-rpc, or --event-contract only for non-standard environments.scripts/regression/shasta/.Use when cutting a versioned raiko2 ZK/runtime source release from this repository, including git tag creation, publishing the runtime image, exporting guest digests, and creating a GitHub Release with release notes and manifest artifacts. TEE provider metadata is a separate optional flow. Use when the workflow must stop before deployment or register-image --apply.
Use when building and publishing a raiko2 runtime image from this repository with just release-image or xtask release-image. Use when Codex must keep the workflow limited to guest ELF refresh, image build/push, digest capture, and optional register-image checks, without any kubectl, GKE, or rollout steps.
Use when integrating a third-party remote proving service such as TDX, SGX, or another TEE-backed provider into raiko2, especially when the provider must implement the raiko2 remote prover protocol, pass black-box conformance, bootstrap or init correctly, and clear API regression on both proposal and aggregate proving.