Record the thought process and scientific justification behind a design or implementation decision, structured for future use in writing a scientific article.
Record the scientific reasoning, trade-offs, and justification behind a design or implementation decision. The output is structured so it can later feed into a methods section, supplementary material, or design-rationale appendix of a scientific paper.
The user will invoke this skill when a decision has just been made (or is being made) that involves meaningful scientific, algorithmic, or architectural reasoning. The user may provide a topic as an argument (e.g., /record-decision conformer deduplication strategy), or you may need to infer it from the current conversation context.
Gather context: Review the current conversation for the decision being discussed. If the conversation doesn't contain enough context, ask the user concise clarifying questions. Focus on:
Draft the record: Write a structured decision record using the template below. Show the draft to the user for review before saving.
Save the record: Write the final record to docs/decisions/NNNN-<slug>.md, where NNNN is the next sequential number (check existing files in docs/decisions/). Create the docs/decisions/ directory if it doesn't exist.
Update the index: Append a one-line entry to docs/decisions/INDEX.md (create it if missing). Format: | NNNN | Title | YYYY-MM-DD | Status |.
# DR-NNNN: <Title>
**Date:** YYYY-MM-DD
**Status:** Accepted | Proposed | Superseded by DR-XXXX
**Authors:** <who participated in the decision>
## Context
_What situation or requirement prompted this decision? What is the scientific or engineering problem being addressed? Include references to relevant literature, standards, or domain conventions where applicable._
## Considered Alternatives
### Alternative A: <name>
- **Description:** ...
- **Pros:** ...
- **Cons / why rejected:** ...
### Alternative B: <name>
- **Description:** ...
- **Pros:** ...
- **Cons / why rejected:** ...
_(repeat as needed)_
## Decision
_State the chosen approach clearly and concisely._
## Scientific Justification
_Why is this the right choice from a scientific perspective? Reference domain knowledge, computational chemistry conventions, thermodynamic/kinetic theory, data integrity requirements, or relevant literature. This section should be written in a style suitable for adaptation into a methods or design-rationale section of a paper._
## Implementation Notes
_How does this decision manifest in the codebase? Reference specific modules, tables, schemas, or workflows affected. Keep brief — the code is the authoritative source._
## Limitations & Future Work
_Known limitations of this approach. Conditions under which this decision should be revisited. Potential improvements deferred for now._
## References
_Citations, DOIs, URLs, or pointers to related decision records._