Freelancers spend 15+ hours/week finding and applying to contracts. This pipeline automates the hunt-to-proposal workflow: scan job feeds, score opportunities, generate tailored proposals, and test for prompt injection vulnerabilities.
| Metric | Value |
|---|---|
| Tests | 310+ passing |
| Agents | 4-agent proposal pipeline (Prospecting, Credential, Proposal, Engagement) |
| Job Scoring | 105-point rubric, 60+ threshold triggers proposals |
| Injection Detection | 60+ patterns across 8 MITRE ATLAS threat categories |
| LinkedIn Scoring | 0-100 prospect scoring with batch processing |
| CLI Overhead | Click-based command parsing |
| Persistence | SQLite for scanner results and tracking |
- Too much time prospecting -- Job scanner monitors Upwork RSS feeds and scores every listing on a 105-point rubric, surfacing only high-value opportunities
- Generic proposals lose -- A 4-agent pipeline generates tailored proposals that match your credentials to each job's requirements
- LLM apps ship with injection vulnerabilities -- Built-in prompt injection testing suite detects 60+ attack patterns across 8 threat categories
- Service 4: Multi-Agent Workflows (Agentic AI Systems)
- Service 6: AI-Powered Personal and Business Automation
- Service 10: Predictive Analytics and Lead Scoring
- Vanderbilt ChatGPT Personal Automation
- IBM Generative AI Engineering with PyTorch, LangChain & Hugging Face
- IBM RAG and Agentic AI Professional Certificate
- Duke University LLMOps Specialization
- Google Digital Marketing & E-commerce Certificate
graph LR
RSS[RSS Feeds] --> Scanner[Job Scanner<br/>Vollna]
Scanner --> Scoring[Scoring Engine<br/>105-pt rubric]
Scoring --> Pipeline[Proposal Pipeline]
Pipeline --> Tracking[Tracking]
Pipeline --> A1[Prospecting Agent]
Pipeline --> A2[Credential Sync Agent]
Pipeline --> A3[Proposal Architect Agent]
Pipeline --> A4[Engagement Agent]
subgraph Security Testing
Injection[Prompt Injection Suite]
Patterns[60+ Detection Patterns]
MITRE[MITRE ATLAS Mapping]
end
subgraph LinkedIn Engine
Scorer[Prospect Scorer 0-100]
Templates[Template Engine]
Batch[Batch Generator + Analytics]
end
ASCII version
RSS Feeds ──> Job Scanner ──> Scoring Engine ──> Proposal Pipeline ──> Tracking
(Vollna) (105-pt rubric) |
|
4-Agent System:
1. Prospecting Agent
2. Credential Sync Agent
3. Proposal Architect Agent
4. Engagement Agent
Security Testing: LinkedIn Engine:
- Prompt Injection Suite 1. Prospect Scorer (0-100)
- 60+ detection patterns 2. Template Engine (3 hook types)
- MITRE ATLAS mapping 3. Batch Generator + Analytics
- Risk scoring + hardening
git clone https://github.com/ChunkyTortoise/Revenue-Sprint.git
cd Revenue-Sprint
bash scripts/setup.sh
# Demo mode — no API keys needed, uses sample data
python scripts/run_pipeline.py --demo
# Run the test suite
pytest tests/ -vcp .env.example .env
# Add your ANTHROPIC_API_KEY
python scripts/run_pipeline.py \
--feed-url scanner/tests/sample_feed.xml \
--dry-runMonitors Upwork RSS feeds (via Vollna), scores listings, stores results in SQLite, and dispatches notifications.
| Criterion | Points |
|---|---|
| RAG/vector/embedding keywords | +30 |
| Agentic/orchestration keywords | +30 |
| Budget >= $5,000 | +25 |
| Client rating >= 4.8 | +10 |
| Posted < 24 hours ago | +10 |
Scores of 60+ trigger proposal generation. Notifications go to console, Slack, and macOS notification center.
Four agents chained end-to-end:
- Prospecting Agent -- Scores job-credential fit (0-100)
- Credential Sync Agent -- Matches job against certifications and portfolio projects
- Proposal Architect Agent -- Generates proposals with configurable angles (ROI-Focused, Technical Deep-Dive, Speed-to-Value)
- Engagement Agent -- Classifies post-proposal client messages and generates strategic responses
All agents fall back to keyword/regex logic when ANTHROPIC_API_KEY is unset.
Detection engine for AI/LLM prompt injection attacks. 60+ detection patterns across 8 threat categories, MITRE ATLAS mapping, severity-based risk scoring, and a hardening advisor.
Automated outreach: prospect scoring (0-100), template generation (connection requests, DMs, follow-up sequences), batch processing, and funnel analytics. Standard library only.
| Layer | Technology |
|---|---|
| Pipeline | Python, Click CLI |
| Agents | Claude AI (keyword fallback without API key) |
| Scanner | RSS parsing, SQLite, multi-channel notifications |
| Security | Regex detection engine, MITRE ATLAS mapping |
| Testing | pytest (310+ tests, all run without API key) |
Revenue-Sprint/
├── portfolio-rag-core/ # 4-agent proposal pipeline
│ └── src/agents/ # Prospecting, Credential, Proposal, Engagement
├── scanner/ # Upwork job scanner + scoring
├── product_1_launch_kit/ # Prompt Injection Testing Suite
├── product_2_launch_kit/ # RAG Cost Optimization Toolkit
├── product_3_launch_kit/ # Multi-Agent Orchestration Kit
├── linkedin_engine/ # LinkedIn outreach automation
├── security_suite/ # Fuzzer + benchmark (250+ cases)
├── scripts/ # setup.sh, verify.py, run_pipeline.py
├── benchmarks/ # Performance benchmarks (CLI, revenue calc)
├── tests/ # 310+ tests
└── requirements.txt
| ADR | Title | Status |
|---|---|---|
| ADR-0001 | Click CLI Over argparse | Accepted |
| ADR-0002 | SQLite for Persistence | Accepted |
| ADR-0003 | AI Analysis Integration | Accepted |
See BENCHMARKS.md for performance methodology and results. Run locally:
python benchmarks/run_all.pypytest tests/ -v # All 310+ tests
pytest tests/ -m unit # Fast isolated tests
pytest tests/ -m integration # SQLite + multi-component
pytest tests/test_product1.py # Injection detection
pytest tests/test_agents.py # Proposal pipeline
pytest tests/test_scanner.py # Job scannerAll tests run without an API key.
See CHANGELOG.md for release history.
- EnterpriseHub -- Real estate AI platform with BI dashboards and CRM integration
- jorge_real_estate_bots -- Three-bot lead qualification system (Lead, Buyer, Seller)
- ai-orchestrator -- AgentForge: unified async LLM interface (Claude, Gemini, OpenAI, Perplexity)
- insight-engine -- Upload CSV/Excel, get instant dashboards, predictive models, and reports
- docqa-engine -- RAG document Q&A with hybrid retrieval and prompt engineering lab
- scrape-and-serve -- Web scraping, price monitoring, Excel-to-web apps, and SEO tools
- Portfolio -- Project showcase and services
MIT