Extract architectural decisions from conversations into MADR-compliant documents.
npx skills add https://github.com/arthrod/conejo-skills --skill adr-arthrod
Please help me install this Agent Skill. Skill: adr Source: https://github.com/arthrod/conejo-skills/tree/main/skills/adr Command: npx skills add https://github.com/arthrod/conejo-skills --skill adr-arthrod
This skill solves the problem of fragmented or lost architectural knowledge by providing a structured, repeatable pipeline to extract, format, and verify Architectural Decision Records (ADRs).
Use the adr skill to extract architectural decisions from the current conversation and format them into a new MADR document.
High-intent search queries and answers about installing and using this skill.
To extract architectural decisions from conversation transcripts, this Skill mines raw conversational context like PR discussions and automatically identifies architectural choices, trade-offs, and rationale for formatting.
MADR compliance for architectural decision records uses the Markdown Architectural Decision Record standard to format extracted technical choices. This ensures consistent, structured documentation of trade-offs and rationale across the engineering team.
You verify the quality of architectural decision records by enforcing the E.C.A.D.R. Definition of Done. This ensures every generated record is Explicit, Comprehensive, Actionable, Documented, and Reviewable for stakeholders.
Yes, you need Python installed to generate MADR documents from PR discussions. Python is required specifically for sequence management during the extraction and formatting pipeline of the architectural decision records.
The best way to standardize architectural documentation across engineering teams is using a structured pipeline that converts raw discussions into formal MADR-compliant records. This prevents fragmented knowledge and enforces a repeatable formatting standard.
A limitation when mining architectural decisions from conversational context is the strict dependency on Python for sequence management. Additionally, all generated records must pass the E.C.A.D.R. verification criteria to be considered complete.
alivecontextExtract Decisions, Tasks, People, Insights, Quotes, and Connections from transcripts, emails, and PDFs.
cosmicstack-labsTurn scattered internal documentation into a structured, searchable wiki.
ccfDocument architectural decisions with a MADR-style template and lifecycle.
carl10086Standardize architectural decision records and project documentation with structured Markdown templates.
