This repository contains the source code for gitforensics.org, a free guide on using Git for evidence preservation and documentation.
Git Forensics teaches how to leverage Git's cryptographic and distributed properties for creating tamper-evident documentation. The guide covers workplace applications, protection strategies, and forensic techniques.
This is a static website built with vanilla HTML, CSS, and JavaScript.
The book content is © 2025 Caia Tech. Free for personal, educational, and non-commercial use.
See the book's copyright notice for full terms.
Please use the ideas here to build amazing tools of the future.
Created by Caia Tech
Git Forensics as the Solution----------------------------------------------------------------
Real-Time AI Verification
Every AI interaction automatically:
- Creates cryptographic commits of all outputs
- Generates distributed timestamps across multiple systems
- Builds tamper-evident chains of conversation
- Provides mathematical proof of authenticity
Implementation Examples
Medical AI Systems
- Diagnosis and treatment recommendations Git-committed in real-time
- Immutable patient interaction logs
- Legally admissible evidence of AI medical advice
Legal AI Assistants
- Every legal opinion cryptographically signed
- Timestamped advice that can't be retroactively altered
- Protection for both lawyers and clients
Financial AI Trading
- Algorithmic decisions recorded immutably
- Regulatory compliance through transparent audit trails
- Proof of AI reasoning at time of trade
Customer Service AI
- Every promise or commitment permanently recorded
- Protection against "we never said that" disputes
- Corporate accountability for AI representatives
The Network Effect
When AI systems adopt Git Forensics:
- Each AI output becomes independently verifiable
- Multiple systems create interlocking verification
- Tampering requires compromising entire networks
- Truth becomes mathematically provable
Beyond Individual Protection
This isn't just about protecting users from AI deception. It's about:
- Enabling AI systems to build trust through transparency
- Creating legal frameworks for AI accountability
- Establishing cryptographic standards for AI evidence
- Building the verification backbone for the AI age
Git Forensics transforms AI from a black box of deniable outputs into a transparent system of cryptographically verified truth. In an era where reality itself is under assault by synthetic media, this method provides the mathematical certainty courts, corporations, and citizens desperately need.
Git Forensics becomes exponentially more powerful through verification networks - multiple independent parties who cannot collude to lie.
Example Network:
- Your organization commits evidence
- Regulator nodes timestamp it
- Competitor nodes verify (mutual watching)
- Academic institutions archive
- International bodies witness
No single entity can forge consensus across hostile witnesses.
Applications:
- Financial markets (exchanges + regulators + competitors)
- Healthcare (hospitals + insurers + oversight boards)
- Supply chains (manufacturers + distributors + customs)
- AI/ML systems (platforms + researchers + watchdogs)
The principle: Truth emerges from parties who distrust each other but trust the math.
Git Forensics works equally well with encrypted data. You can have both verification and privacy.
How it works: Encrypt sensitive content, commit the encrypted blob + hash, distribute widely, selectively share decryption keys.
What's preserved: Timestamps, tamper-evidence, distribution, and your control over disclosure.
Temporal + Git Forensics: The Next Evolution Automated Evidence Collection with Immutable Workflow History
While Git forensics provides cryptographic proof of WHAT was documented and WHEN it was committed, Temporal adds another critical layer: HOW evidence was collected and processed.
Key Synergies:
- Workflow History as Legal Evidence
- Every automated collection action is permanently recorded
- Timestamps for each step in the evidence gathering process
- Immutable audit trail of all workflow executions
- Proves evidence was collected contemporaneously, not fabricated later
- Durable Execution for Continuous Documentation
- Workflows survive system failures, attacks, and interruptions
- Long-running workflows can document patterns over months/years
- Automatic retries with full history preservation
- Creates unbreakable chain of custody
-
Dynamic Response to Evolving Threats // Example: Adaptive evidence collection async function monitorAndDocument() { while (true) { const threat = await detectNewThreat(); // Dynamically spawn appropriate documentation workflow await startChild(
document${threat.type}Workflow); // Commit findings to Git with cryptographic proof await commitToGitRepo(threat); } } -
Query System for Legal Discovery
- "Show all evidence collected between date X and Y"
- "Prove when this specific document was first observed"
- "Demonstrate the sequence of events leading to this commit"
- All queries return immutable, timestamped results
Why This Matters:
- Git proves the evidence is authentic
- Temporal proves the collection process was legitimate
- Together: Unassailable documentation system
- Courts can verify both the evidence AND the methodology
Use Cases:
- Automated monitoring of changing documents/databases
- Real-time documentation of ongoing violations
- Proof of contemporaneous evidence collection
- Protection against claims of fabrication or backdating The combination creates a legally robust system where every piece of evidence has both cryptographic verification (Git) and process verification (Temporal).
note: Also through combination of temporal in git, was able to create https://github.com/Caia-Tech/volcano-llm
Note: Temporal is open source and can be self-hosted, maintaining the same independence principles as Git forensics.
"This developer flies solo" 🛩️