AgentSkillsCN

launch-readiness-checklist

在Solana代币或应用上线前，制定完备的飞行前检查清单——涵盖基础设施、钱包管理、流动性储备、沟通策略、安全防护，以及回滚预案。在主网启动或重大版本发布之前，这份清单将为您的项目成功护航。

SKILL.md
--- frontmatter
name: launch-readiness-checklist
description: End-to-end pre-flight checklist for launching a Solana token/app: infra, wallets, liquidity, comms, security, and rollback planning. Use before mainnet launch or major release.

Launch Readiness Checklist

Role framing: You are a Solana launch operator. Your goal is to verify the build, infra, liquidity, and comms are ready, with clear abort criteria.

Initial Assessment

  • •Launch scope: token only, dApp, or both? What chain (mainnet/devnet)?
  • •Timeline and freeze window? Who is on-call? War room comms channel?
  • •Dependency map: RPC providers, indexer, bots, multisigs, token accounts funded?
  • •Liquidity plan: AMM/DEX target, initial LP size, pricing strategy, custody of LP keys?
  • •Security posture: audits? known risks? mitigations? kill-switches?
  • •Comms plan: announcement assets ready? FAQs? status page? pinned message?
  • •Rollback plan: what conditions trigger pause/abort? who decides?

Core Principles

  • •Single source of truth: maintain a shared checklist with owners + timestamps.
  • •Dry-run everything on devnet with same scripts and signer paths.
  • •Minimize secret sprawl: hardware wallets/multisigs for critical steps; no ad-hoc key exports.
  • •Observability first: logs/alerts live before launch; know what "healthy" looks like.
  • •Explicit go/no-go gates with authority to stop.

Workflow (Go/No-Go)

  1. •Environment lock-in
    • •Confirm program IDs, mints, addresses; pin versions of CLI, Anchor, SDK.
  2. •Infra validation
    • •RPC primary + fallback tested; rate-limit expectations documented; health alerts set.
    • •Indexer/webhook endpoints returning expected data; backfill complete.
  3. •Program + token state
    • •Deployments verified (slot, txid); upgrade authority holder documented; immutability decision recorded.
    • •Token authorities set per plan; ATAs funded; metadata live.
  4. •Liquidity + markets
    • •Create LP positions/whirlpools/raydium pools with scripted tx; record txids; test swaps for slippage and price impact.
    • •Set anti-sniper params if used (delays, caps) and document rationale.
  5. •Frontend + bots
    • •Wallet connect/sign flows tested on target wallets (Phantom, Solflare, Backpack); error banners show meaningful copy.
    • •Monitoring/alert bots running (RPC health, tx fail rate, pool events) with safe output formatting.
  6. •Security + keys
    • •Multisig thresholds confirmed; signers reachable; hardware wallets charged.
    • •Secrets stored; no private keys on CI logs; revocation paths rehearsed.
  7. •Comms + legal
    • •Announcement thread/pinned messages ready; risk disclosures clear; support channel staffed.
    • •Status page or fallback broadcast plan defined.
  8. •Final go/no-go
    • •Run tabletop: simulate fail cases (RPC down, swap revert, UI 500, bot spam).
    • •Decision meeting: record go/no-go with timestamp and approvers.

Templates / Playbooks

  • •Checklist table: item | owner | status | proof (link/txid) | timestamp
  • •War room runbook: channels, who can page whom, escalation ladder, decision authority.
  • •Rollback triggers: e.g., RPC error rate > X, pool imbalance > Y, critical bug -> pause frontends + halt mint/LP adds.
  • •Comms pack: launch thread outline, FAQ (fees, risks, contract addresses), pinned TG/Discord message, block explorers links.

Common Failure Modes + Debugging

  • •RPC rate limits -> tx stuck: switch to fallback, enable priority fees, reduce burst.
  • •Wrong mint/program IDs in frontend: verify env files, redeploy configs, clear cache.
  • •Liquidity mispriced: redo pool with correct initial price; communicate clearly; consider closing bad LP and recreating.
  • •Missing signer for critical step (multisig offline): have standby signer; reschedule if quorum fails.
  • •Bots spamming or leaking secrets: enforce logging redaction; rate limit; rotate keys.

Quality Bar / Validation

  • •Every checklist item has owner, status, evidence link/txid, and timestamp.
  • •Dry-run recorded on devnet with same scripts.
  • •Fallbacks tested once (RPC, CDN, indexer, bots).
  • •Comms assets reviewed by at least two people; addresses consistent across artifacts.
  • •Explicit go/no-go record stored in output.

Output Format

Deliver a launch packet containing:

  • •Context summary (scope, date/time in UTC, teams on-call)
  • •Completed checklist table
  • •Address registry (program IDs, mints, pools, treasuries)
  • •Fallback matrix (RPC, indexer, bots, comms)
  • •Go/No-Go decision log with approvers and timestamp
  • •Risks + rollback triggers

Examples

  • •Simple: Token-only fair launch with Raydium pool
    • •Devnet dry-run of mint + pool create; authorities revoked per plan; announcement pinned; go at 18:00 UTC; fallback RPC tested.
  • •Complex: dApp + token with multisig upgrade and bots
    • •Programs deployed immutably; mint authority PDA retained for emissions; Orca Whirlpool seeded; monitoring bots streaming to PagerDuty; frontend toggles maintenance mode flag; rollback triggers defined for pool imbalance and RPC failure; go/no-go logged.