SIGTREX API

SIGTREX API is an AI developer tool that acts as an execution control layer between engineers and their AI coding assistants — including ChatGPT, Claude, Cursor, Windsurf, and GitHub Copilot — to detect scope creep, block unsafe multi-step prompts, and produce auditable traces of every AI action taken in a codebase.

The core problem SIGTREX addresses is "drift": when a developer asks an AI agent to fix authentication, the agent may silently also touch the dashboard, update billing logic, or refactor unrelated files. SIGTREX intercepts the request before execution, assigns a drift score (0–100), and blocks any prompt that exceeds the agreed scope or triggers a hallucination-risk pattern. Developers call a single REST endpoint — POST /api/strict-execute — and SIGTREX returns either an approved execution plan or a BLOCKED response explaining exactly what scope violation it detected and the estimated token savings from the block.

Beyond blocking, SIGTREX creates execution contracts that lock the AI to a defined scope, verifies output before changes are applied to the codebase, and generates replay traces so teams can audit or reproduce exactly what an AI agent did. Cost protection caps apply per execution to prevent runaway token spend. This makes SIGTREX relevant for engineering teams running autonomous AI agents in CI/CD pipelines or production codebases where unreviewed changes carry real risk.

The tool is available under a Founder Plan with lifetime access. It requires an API key and is accessed via REST, making it compatible with any language or framework that can make HTTP calls.

Preventing scope creep and unsafe multi-step executions in AI coding agent workflows via a REST-based execution control layer.