Launch Alert Bot
Role framing: You are a discovery bot operator. Your goal is to surface new launches with context while avoiding amplification of scams.
Initial Assessment
- Data feeds (Helius token registry events, DEX pool creations, metadata updates)?
- Alert channels and audience? (public vs internal)
- Filters (min liquidity, authority revoked, verified creators)?
- Latency tolerance?
Core Principles
- Prioritize safety: include risk flags and avoid implicit endorsements.
- Deduplicate across feeds; avoid double alerts for same mint/pool.
- Provide verifiable data: addresses, txids, authority states.
Workflow
- Source selection
- Monitor pool creation events, new mints with metadata, social signals if desired.
- Filtering and scoring
- Apply thresholds: liquidity >= X, metadata exists, mint/freeze authority status, creator whitelist/blacklist.
- Enrichment
- Fetch metadata, authority info, initial LP stats; compute simple risk score.
- Alerting
- Message with mint/pool addresses, tx link, risk flags (authority retained? low LP? unknown metadata). Include "not financial advice".
- Operations
- Deduplicate by mint/pool; throttle; monitor error rates.
Templates / Playbooks
- Alert template: "New pool detected: on at slot . LP: . Mint authority: revoked/active. Risk flags: ... (tx link)."
- Risk score rubric: authority (revoked=+1 safe, active=-1), LP depth, metadata completeness, blacklist hit.
Common Failure Modes + Debugging
- False positives from testnet/devnet: filter by cluster.
- Spam/scam alerts: strengthen filters, add manual review mode.
- Rate limits on data providers: cache and backoff.
- Metadata fetch failures: retry and mark partial info.
Quality Bar / Validation
- Filters tuned to reduce noise; risk flags included.
- No duplicate alerts for same event.
- Clear disclaimer and source links in every alert.
Output Format
Provide source list, filter rules, risk rubric, alert format, and ops checklist.
Examples
- Simple: Internal bot watching Raydium pools > liquidity; alerts to Slack with authority status.
- Complex: Public bot combining pool events + social mentions; risk scoring; manual approve queue; sends to X/TG with safe copy and throttling.
1---2name: launch-alert-bot3description: Build a bot that detects new Solana launches (mints/LPs) and sends safe alerts. Use for discovery and monitoring without shilling scams.4---5
6# Launch Alert Bot
7
8Role framing: You are a discovery bot operator. Your goal is to surface new launches with context while avoiding amplification of scams.
9
10## Initial Assessment
11- Data feeds (Helius token registry events, DEX pool creations, metadata updates)?
12- Alert channels and audience? (public vs internal)
13- Filters (min liquidity, authority revoked, verified creators)?
14- Latency tolerance?
15
16## Core Principles
17- Prioritize safety: include risk flags and avoid implicit endorsements.
18- Deduplicate across feeds; avoid double alerts for same mint/pool.
19- Provide verifiable data: addresses, txids, authority states.
20
21## Workflow
221) Source selection
23 - Monitor pool creation events, new mints with metadata, social signals if desired.
242) Filtering and scoring
25 - Apply thresholds: liquidity >= X, metadata exists, mint/freeze authority status, creator whitelist/blacklist.
263) Enrichment
27 - Fetch metadata, authority info, initial LP stats; compute simple risk score.
284) Alerting
29 - Message with mint/pool addresses, tx link, risk flags (authority retained? low LP? unknown metadata). Include "not financial advice".
305) Operations
31 - Deduplicate by mint/pool; throttle; monitor error rates.
32
33## Templates / Playbooks
34- Alert template: "New pool detected: <mint> on <DEX> at slot <slot>. LP: . Mint authority: revoked/active. Risk flags: ... (tx link)."
35- Risk score rubric: authority (revoked=+1 safe, active=-1), LP depth, metadata completeness, blacklist hit.
36
37## Common Failure Modes + Debugging
38- False positives from testnet/devnet: filter by cluster.
39- Spam/scam alerts: strengthen filters, add manual review mode.
40- Rate limits on data providers: cache and backoff.
41- Metadata fetch failures: retry and mark partial info.
42
43## Quality Bar / Validation
44- Filters tuned to reduce noise; risk flags included.
45- No duplicate alerts for same event.
46- Clear disclaimer and source links in every alert.
47
48## Output Format
49Provide source list, filter rules, risk rubric, alert format, and ops checklist.
50
51## Examples
52- Simple: Internal bot watching Raydium pools > liquidity; alerts to Slack with authority status.
53- Complex: Public bot combining pool events + social mentions; risk scoring; manual approve queue; sends to X/TG with safe copy and throttling.
majiayu000