What problem does it solve?
Reverse-engineering front-end JavaScript signature parameters (like h5st, a-bogus, falcon) is usually ad-hoc and error-prone. This Skill standardizes the entire workflow—page observation, runtime hook sampling, local Node reproduction, environment patching, and deobfuscation—so every step produces traceable evidence artifacts instead of guesswork.
Core Features & Use Cases
- Five-Stage Workflow: Enforces an Observe → Capture → Rebuild → Patch → DeepDive pipeline with explicit rules like observe-first, hook-preferred, and breakpoint-last.
- Task-Driven Tool Routing: Directs the AI to the right MCP tools (start_reverse_task, orchestrate_reverse_task, run_reverse_agent, export_portable_bundle) based on user intent, with a structured response-reading protocol.
- Environment Patching Discipline: Drives Node environment patching via proxy env logs, first-divergence tracking, and minimal causal-unit patches rather than blind browser simulation.
- Use Case: When asked to reproduce a site's request signature in Node, the Skill first locates the target request and script in the browser, hooks fetch/xhr for runtime samples, exports a rebuild bundle, then iteratively patches the environment until the signature verifies.
Quick Start
Use the jsreverser-mcp-playbook to reverse the signature parameter of the target site's API, starting a new reverse task and following the observe-first workflow.
