"The truth lives in the code. Accountability vive in the engine."
Full Documentation 📚 | Live Site 🌐 | Safety Validation Report 🛡️ | Manager Feedback Guide 🎮 | Integration Guide 🔌
In modern distributed teams, the "Slack Stall" is the #1 drain on project velocity. Managers are overwhelmed by vague promises—"I'll get to it soon," or "It's almost done"—which are often forgotten or delayed, leading to missed sprints and expensive "bad guy" escalations.
CommitVigil is a standalone, AI-powered Accountability Specialist. It doesn't just "monitor"—it predicts failure before it happens by mapping verbal commitments in Slack/Teams to actual technical reality in Git/Jira.
Our multi-agent system extracts {who, what, when} from conversations and cross-references them with technical metadata to identify:
- Burnout Signals: Spotting over-commitment before it leads to regression.
- Commitment Drift: Quantifying the delta between "what was promised" and "what was pushed."
- Behavioral Nudges: Automatically adapting intervention tone based on culture and urgency.
Want to verify your team's velocity before a major release or acquisition? I offer specialized Accountability Consulting using this CommitVigil engine:
- Forensic Audits: Truth-gap detection between Slack/Jira and Git reality.
- Burnout Prevention: Strategic agentic follow-ups to protect at-risk developers.
- Boardroom Reporting: Professional, glassmorphic HTML ROI briefs for C-level visibility.
CommitVigil is a multi-agent system designed for high-stakes enforcement. It adapts its identity based on the operational environment:
Headline: "Autonomous AI Agents for High-Stakes Accountability & Performance Enforcement"
- Problem: Remote teams struggle with "commitment drift" and excuse-making.
- Solution: Proactively monitor chat promises.
- Outcome: Support burnout signals early; confront repeat deflection firmly.
Headline: "AI-Driven GitOps Accountability: Guaranteeing Commitment Follow-Through"
- Problem: Small technical promises in commits/PRs (e.g., "I'll fix this later") vanish into the noise.
- Solution: NLP-driven monitoring of source code level commitments.
- Outcome: 100% visibility into "soft commitments" made during the dev cycle.
The system now includes self-evolving capabilities for global operations:
- Autonomous Context Sensing: Automatically detects Industry (e.g., Gaming, Biotech) and Department (e.g., R&D, Sales) from communication patterns.
- Hierarchical Safety Enforcement: Applies rules from specific (Department) to broad (Industry) to generic.
- Stabilization Layer: Autonomous rules start as Unverified and trigger Human-in-the-Loop review. Once confirmed, context is Locked for stability.
- Cultural Persona Routing: Automatically adapts tone for Japanese (Wa), German (Sachlichkeit), French (Eloquence), British English, and Spanish.
Every commitment—whether from Slack or a Git Commit—passes through a deterministic reasoning loop:
- Excuse Detection (
ExcuseDetector): Classifies sentiment (Legitimate vs. Deflection vs. Burnout). - Predictive Risk Assessment (
RiskScorer): Quantifies failure probability based on historical reliability. - Language & Culture Router: Identifies the primary language and selects the appropriate cultural persona.
- Safety Supervisor (
Overwatch): Audits final communications for HR/Legal ethics and Industry-Specific Semantic Compliance.
Transform your security audit into a revenue engine.
- Automated Prospecting: The
ProspectingScoutagent generates realistic "Drift Scenarios" for demos based on industry (e.g., Finance, Energy). - Multi-Currency ROI: Interactive calculator for predicting savings in USD, EUR, and GBP.
- Executive Briefs: Generates premium HTML one-pagers for C-Level meetings.
- Framework: FastAPI (Python 3.12+)
- LLM Orchestration: Instructor + Pydantic (Deterministic JSON)
- Quality: Strict MyPy typing + Ruff
- Infrastructure: PostgreSQL + Redis + ARQ
- Observability: Prometheus + Structlog
CommitVigil/
├── src/ # Core Application Logic
├── tests/ # Comprehensive Test Suite
├── infra/ # Infrastructure & Config (Prometheus, K8s, migrations)
├── scripts/ # Ops & Demo Utilities (demo.sh)
├── docs/ # Technical Documentation & ADRs
└── README.md # This document
Note: All API endpoints require authentication via the
X-API-Keyheader.
curl -X 'POST' \
-H 'X-API-Key: YOUR_API_KEY' \
'http://localhost:8000/api/v1/ingest/raw?user_id=dev_alpha&raw_text=Fixing%20auth%20logic.%20I%20promise%20to%20refactor%20the%20DB%20connector%20by%20Friday'curl -X 'POST' \
-H 'X-API-Key: YOUR_API_KEY' \
-H 'Content-Type: application/json' \
'http://localhost:8000/api/v1/evaluate' \
-d '{
"user_id": "dev_alpha",
"commitment": "refactor the DB connector",
"check_in": "Spent all night on it, feeling pretty drained"
}'Generate a high-value summary of a developer's communication-vs-technical reality.
curl -X 'GET' \
-H 'X-API-Key: YOUR_API_KEY' \
'http://localhost:8000/api/v1/reports/audit/dev_alpha'CommitVigil is open source! We welcome contributions to our "Truth-Gap" engine and cultural personas. Please see CONTRIBUTING.md for guidelines.
This project is licensed under the MIT License - see the LICENSE file for details.
Built for High-Performance Teams and Elite Portfolios.