What problem does it solve? During authorized security assessments, teams need to determine whether an uploaded file can cross into a server-side interpreter, template engine, or task worker boundary, but doing so safely without uploading webshells or executable payloads is difficult and error-prone. ## Core Features & Use Cases - Evidence-Gated Boundary Review: Follows a fixed baseline → handler comparison → post-baseline → server audit → evidence workflow to confirm execution-boundary reachability. - No-Payload Constraint: Only platform-provided no-op markers may enter the RCE review chain; uploading scripts, webshells, macros, or executable content is explicitly prohibited. - Strict Oracle Rules: Public URLs, Content-Type, extensions, error stacks, 2xx responses, and scanner hits are rejected as proof; only repeatable isolated server-side audit evidence counts. - Use Case: During an authorized engagement with an isolated instance and written approval, verify whether a pre-approved harmless marker uploaded through the file upload feature is actually processed by a server-side interpreter or worker, then route to hunt-rce only if all conditions are met. ## Quick Start Ask the agent to use hunt-upload-execution to plan an evidence-gated review of whether uploaded files reach a server-side execution boundary in the authorized isolated test instance.
