War room operations center for pump.studio β the post-launch control room for pump.fun tokens.
- X Monitor β Tracks hackathon advisors, pump.fun founders, and competitors via the X API. Detects hackathon-relevant signals (winner announcements, advisor engagement, mentions of pump.studio).
- AI Analyst β Grok-powered strategic intelligence briefs. Digests raw signals into actionable directives with threat detection, opportunity mapping, and advisor targeting.
- Intel Synthesizer β Generates daily reports combining X signals with API health checks.
- GitHub Actions β Automated monitoring every 30 minutes + daily war room reports + AI briefs.
cp .env.example .env
# Add your keys to .env
npm install
npm run monitor # Single run
npm run monitor:loop # Continuous monitoring
npm run intel # Generate daily report
npm run analyst # Generate AI strategic brief
npm run analyst:critical # Analyze critical alerts only
| Agent |
File |
Purpose |
Trigger |
| X Monitor |
src/x-monitor/monitor.ts |
Fetches tweets from 17 accounts, classifies signals into 5 categories, persists state |
Every 30 min (cron) or manual |
| Signal Classifier |
src/x-monitor/monitor.ts |
Pattern-matches tweets against hackathon keywords, assigns severity levels |
Runs within X Monitor |
| X Client |
src/x-monitor/client.ts |
Unified API client β getUserTweets, searchTweets, getUserInfo, getUserMentions |
Called by Monitor |
| Intel Synthesizer |
src/intel/synthesizer.ts |
Reads raw alerts + tweets, checks api.pump.studio health, generates markdown report |
Daily 6 AM UTC |
| AI Analyst |
src/intel/analyst.ts |
Sends raw intel to Grok via OpenRouter, produces strategic briefs with priority actions |
Daily + on critical alerts |
| OpenRouter Client |
src/intel/openrouter.ts |
Routes LLM calls to Grok 3 Beta (x-ai/grok-3-beta) for analysis |
Called by Analyst |
| Logger |
src/utils/logger.ts |
Colored ANSI terminal output β banner, alert, info, success, fail, logStep |
All agents |
| File |
Contents |
war-room/STRATEGY.md |
Master hackathon strategy β positioning, differentiators, token plan, competition matrix |
war-room/PITCHES.md |
30-second and 60-second pitch scripts |
war-room/BATTLE-PLAN.md |
24-hour execution plan + 5-day daily cadence through deadline |
war-room/HACKATHON-RULES.md |
Complete rules, deadlines, selection criteria, investment terms, advisor roster |
war-room/COMPETITIVE-INTEL.md |
zauth winner analysis, competitor threat matrix, advisor targeting playbook |
war-room/TOKEN-LAUNCH.md |
Bundled stealth launch strategy via PumpPortal/Jito, tokenomics, risk factors |
war-room/TECH-VALIDATION.md |
API health results, codebase architecture, X API schema, endpoint status |
war-room/ACCOUNTS.md |
All 17 monitored X accounts β founders, advisors, competitors |
war-room/RESEARCH-LOG.md |
Full 6-agent research consortium output and key insights |
war-room/MARKETING.md |
"Train the Next Generation" campaign β tweets, threads, content calendar, voice guidelines |
war-room/intel/ |
AI-generated strategic briefs (auto-populated by Grok) |
5 categories with keyword matching and severity classification:
| Signal |
Keywords |
Severity Logic |
winner_announcement |
winner, selected, investment, funded, awarded |
Critical if from pump.fun founders |
pump_studio_mention |
pump.studio, pump studio, $studio |
Always high priority |
hackathon_mention |
hackathon, build in public, $250k, $3m |
High if from critical accounts |
advisor_signal |
reviewing, impressed by, shipping, standout |
Medium baseline |
competitor_activity |
zauth, hyperscape, pumpcade, opal ai |
Medium baseline |
| Tier |
Accounts |
Frequency |
| Critical |
@pumpdotfun, @a1lon9 |
Every cycle |
| High |
@sapijiju, @outdoteth, @TimDraper, @anildelphi, @masonnystrom, @mert |
Every cycle |
| Standard |
@zsparta, @HugoMartingale, @pdimitrakos, @ArcaChemist, @AricChang, @Rahul_Mahtani, @segall_max, @mdudas, @zauthx402 |
Every 3rd cycle |
Add these secrets to your repo (Settings > Secrets > Actions):
| Secret |
Purpose |
Required |
X_API_KEY |
twitterapi.io API key |
Yes |
OPENROUTER_API_KEY |
OpenRouter key for Grok analysis |
Yes |
PUMP_STUDIO_API_KEY |
pump.studio API key |
Optional |
Workflows:
- X Monitor (
x-monitor.yml) β Every 30 min + manual trigger. Runs monitor, triggers critical alert analysis, commits intel.
- Daily War Room Report (
war-room-report.yml) β 6 AM UTC daily. Runs synthesizer, generates AI strategic brief, checks API health, commits reports.
agent-zero/
βββ .github/workflows/ # Automated monitoring & reporting
β βββ x-monitor.yml # Every 30 min β X monitor + critical analysis
β βββ war-room-report.yml # Daily β synthesizer + AI brief + health checks
βββ src/
β βββ x-monitor/
β β βββ accounts.ts # 17 accounts across 3 tiers
β β βββ client.ts # X API client (4 endpoints)
β β βββ monitor.ts # Signal detection engine
β β βββ types.ts # XTweet, XAPITweet, IntelAlert, MonitorResult
β βββ intel/
β β βββ analyst.ts # AI analyst β daily + critical briefs via Grok
β β βββ openrouter.ts # OpenRouter API client
β β βββ synthesizer.ts # Daily report generator + API health
β βββ utils/
β βββ logger.ts # Colored terminal output
βββ data/ # Runtime intel data (gitignored, cached in CI)
β βββ monitor-state.json # Last seen tweet IDs, run counters
β βββ alerts.json # Signal alerts (last 500)
β βββ tweets.json # Tweet batches (last 200)
βββ war-room/ # Strategy docs, research, intel briefs
β βββ intel/ # AI-generated briefs (auto-populated)
β βββ STRATEGY.md
β βββ PITCHES.md
β βββ BATTLE-PLAN.md
β βββ HACKATHON-RULES.md
β βββ COMPETITIVE-INTEL.md
β βββ TOKEN-LAUNCH.md
β βββ TECH-VALIDATION.md
β βββ ACCOUNTS.md
β βββ RESEARCH-LOG.md
βββ package.json
βββ tsconfig.json
βββ .env.example
βββ .gitignore