Research Interrogator
Builds project context and blocks drafting until confidence clears the quality gate.
Open-source research infrastructure
Write research with guardrails.
Nine specialist AI agents, six immutable writing laws, and a traceable workflow that turns your AI code editor into a research writing command center.
9Specialist agents
6Immutable writing laws
3Orchestration layers
90%+Confidence gate
The research team
One agent for every hard part.
Each role came from a real failure mode encountered during thesis research—not from a theoretical workflow.
Research Writer
Drafts traceable prose using research questions and T-C-E-L paragraph structure.
Research Advisor
Challenges research design, argument structure, and defense readiness.
Research Reviewer
Runs an adversarial six-pillar review before work is accepted.
Citation Checker
Verifies references through a DOI chain and flags fabricated citations.
Literature Review
Supports PRISMA-informed search, synthesis, and gap validation.
LaTeX Formatting
Handles venue templates, references, figures, and compilation failures.
Figure Generation
Creates code-first, reproducible research visuals with traceability.
Snapshot Manager
Saves, restores, and compares writing states without requiring Git expertise.
The quality floor
Six laws. No override.
Every paragraph must remain connected to the research design and supported by evidence.
Research-question traceability
No unsupported claims
Verified citation chain
Explicit confidence gates
Domain-aware writing
Reproducible outputs
Quick start
Clone. Configure. Research.
Bring the kit into your AI code editor, complete the project configuration, and let the confidence gate establish a reliable foundation.
git clone https://github.com/Sai21112000/Thesis-Agent-Kit.git
cd Thesis-Agent-Kit
# Open the repository in Cursor or Windsurf
# Start with FIRST-TIMERS-GUIDE.md
# Complete PROJECT-CONFIG.mdBuilt from a defended thesis
Stop fighting your thesis alone.
Open source, domain-agnostic, and designed for accountable research.