What problem does it solve? Research projects accumulate phases, plans, and requirements that become hard to track once a milestone finishes. This Skill closes out a completed research milestone by archiving the roadmap and requirements, generating a research digest, evolving the project state, and tagging the release in git so the next stage starts from a clean, well-documented baseline. ## Core Features & Use Cases - Milestone Archival: Extracts full milestone details into .gpd/milestones/v[X.Y]-ROADMAP.md and v[X.Y]-REQUIREMENTS.md, keeping the active ROADMAP.md constant-size. - Research Digest Generation: Composes a structured RESEARCH-DIGEST.md with narrative arc, key results, methods, convention evolution, figures registry, open questions, and dependency graph for downstream paper writing. - Project Evolution Review: Moves completed requirements to Validated in PROJECT.md, audits key decisions, updates context, and refreshes STATE.md. - Git Integration: Handles branch merging (squash or no-ff), creates annotated version tags, and commits milestone artifacts. - Use Case: After finishing all phases of a v1.0 model derivation milestone, run this Skill to archive the roadmap, generate the research digest, tag v1.0 in git, and prepare handoff notes for the v1.1 validation stage. ## Quick Start Ask the AI to complete milestone v1.0 by archiving the roadmap and requirements, generating the research digest, and tagging the release in git.