Extract settled design decisions and save them as structured markdown crystal files.
npx skills add https://github.com/n0rvyn/indie-toolkit --skill crystallize
Please help me install this Agent Skill. Skill: crystallize Source: https://github.com/n0rvyn/indie-toolkit/tree/main/dev-workflow/skills/crystallize Command: npx skills add https://github.com/n0rvyn/indie-toolkit --skill crystallize
This Skill extracts settled decisions from the current conversation and creates a persistent, machine-readable decision crystal that downstream plan-writing and verification steps can consume, preventing loss of context across sessions.
Run crystallize to extract and save the settled decisions from the current conversation as a decision crystal file.
High-intent search queries and answers about installing and using this skill.
To preserve design decisions after brainstorming, extract settled decisions from your conversation and save them as structured markdown files. This captures constraints, rejected alternatives, and final choices to prevent context loss across sessions.
A decision crystal file is a standardized markdown document used for tracking design choices. It structures Initial Ideas, Discussion Points, Rejected Alternatives, and Decisions with Source Context to ensure traceability and machine-readable knowledge persistence.
Document rejected alternatives and constraints for handoff by scanning the conversation context for decision indicators and persisting them as markdown files. This ensures the next session can verify decisions against the original scope boundaries.
Yes, you can save settled decisions directly to your project documentation directory. The process requires write permission to save dated topic crystal markdown files into the docs folder for downstream plan writing.
The best way to lock in scope boundaries before writing a plan is to extract explicit user decisions and revert events from the current context into a persistent crystal file. This allows the plan-writing step to consume verified constraints.
Yes, the decision tracking process captures inferred items separately from explicit decisions. It scans for high- and medium-signal indicators, marking AI-supplemented items distinctly from direct user decisions within the structured markdown output.
agentflockConduct structured design brainstorming sessions for Myna development.
qte77Extract decisions, rejected alternatives, and patterns from Claude Code plan files.
Ard2pExecute plan tasks sequentially and track progress across sessions.
F0WCUSSave conversation logs to tagged markdown files with automated topic analysis.
