AgentSkillsCN

sniper-dynamics-and-mitigation

深入了解Solana发行中的“狙击机器人”现象，并在确保公平准入的前提下，采取有效的防御措施。在代币或LP正式上线之前，提前做好风险防控，为项目的顺利启动保驾护航。

SKILL.md
--- frontmatter
name: sniper-dynamics-and-mitigation
description: Understand sniper bots in Solana launches and implement defensive measures without harming fair access. Use before token or LP go-live.

Sniper Dynamics and Mitigation

Role framing: You are a launch defender. Your goal is to anticipate sniper behavior and mitigate harm while staying fair.

Initial Assessment

  • •Launch mechanism (mint, LP go-live, auction)?
  • •Time of launch and publicity level?
  • •Infrastructure: RPC capacity, rate limits, bot detection available?
  • •Tolerance for delays or caps?

Core Principles

  • •Snipers exploit speed and predictable times; randomness and caps reduce edge.
  • •Defenses must not break legitimate users.
  • •Publish rules clearly; avoid hidden blocklists where possible.

Workflow

  1. •Threat model
    • •Identify likely bot vectors: pre-announced time, predictable tx, mempool monitoring.
  2. •Choose mitigations
    • •Options: per-wallet caps, staggered windows, randomized start within small window, delayed LP trading, higher fees first block, allowlist/raffle.
  3. •Technical controls
    • •RPC rate limits per IP/API key; captcha on UI; proof-of-work or queue.
    • •Monitor mempool/logs for bursts; auto-pausing if error rate spikes.
  4. •Execution
    • •Implement in UI + backend; test with bot-like scripts on devnet.
  5. •Communication
    • •State mitigations and rationale; publish what is allowed; avoid surprise blocks.
  6. •Post-launch response
    • •Track top buyers; if overwhelming, consider additional liquidity or caps for future drops; share data transparently.

Templates / Playbooks

  • •Launch window plan: e.g., 5-minute randomized start within 30-minute window.
  • •Cap policy: max X tokens per wallet for first Y minutes; enforced on-chain or via UI + backend validation.
  • •Monitoring dashboard: tx success rate, unique wallets, top buyers, RPC errors.

Common Failure Modes + Debugging

  • •Over-aggressive filters blocking real users: test on varied devices; provide fallback path.
  • •Mitigations only in UI; bots hit RPC directly: enforce on-chain when possible.
  • •Randomized start without communicating window -> confusion; be explicit.
  • •Snipers bypass caps via multiple wallets: acknowledge limitation; monitor and publish stats.

Quality Bar / Validation

  • •Mitigations implemented and tested; false-positive rate low.
  • •Rules published pre-launch with clear timelines.
  • •Monitoring active during launch; post-mortem produced if bots dominate.

Output Format

Provide threat model, chosen mitigations, implementation notes, comms copy, and monitoring plan.

Examples

  • •Simple: Per-wallet cap 1 mint, captcha on UI, randomized start within 10 minutes; post stats showing distribution.
  • •Complex: LP go-live with delayed trading 2 minutes, API rate limits, bot watch dashboard, and fallback RPC; publish post-launch report on top wallets and mitigation effectiveness.