| name | term-decision |
| description | Use when selecting terminology and applying consistent replacements across documentation. |
| user-invocable | true |
| disable-model-invocation | true |
Terminology Decision
Overview
Decide contested terms, record rationale, apply replacements, and verify consistency.
Core principle: No global replacement without explicit term decision and verification.
The Process
Step 1: Resolve Decision Scope
- If
$ARGUMENTS exists, focus on that term.
- Otherwise collect candidates from:
- source glossary / terminology pages in extracted markdown
- source index / appendix term lists
term_read.py inconsistency output
term_generate.py missing candidates
- user-flagged terminology
- Create TodoWrite items per term.
Step 2: Prepare Decision Brief
Before presenting a term to the user, check whether the source book already defines or lists it in:
- glossary pages
- index pages
- summary tables / reference pages
If strong source evidence exists, do one first-pass term translation directly:
- add obvious, low-ambiguity terms to
glossary.json
- reserve user discussion only for uncertain, tone-sensitive, or mechanics-ambiguous terms
For each term, present:
- source term
- source glossary/index evidence when available
- current variants and usage locations
- 2-3 candidate translations with short rationale
- related conventions from
style-decisions.json
Step 3: Ask User Decision (Traditional Chinese)
Collect:
- chosen translation
- allowed context-specific variants (if any)
- rationale note
Step 4: Persist Decision
Record in style-decisions.json and update glossary.json:
uv run python scripts/term_edit.py --term "<TERM>" --set-zh "<ZH>" --status approved --mark-term --notes "<REASON>"
Step 5: Apply and Verify
- Preview replacement impact across docs.
- Apply approved replacements.
- Verify:
uv run python scripts/term_read.py --fail-on-forbidden
Progress Sync Contract (Required)
- One TodoWrite item per term decision.
- Mark replacement item complete only after verification passes.
When to Stop and Ask for Help
Stop when:
- term choice changes mechanics interpretation
- user preference conflicts with existing glossary policy
- replacement scope is ambiguous
When to Revisit Earlier Steps
Return to Step 2 when:
- new conflicting evidence appears
- user changes proper noun policy
Red Flags
Never:
- apply unapproved term variants globally
- skip post-edit glossary validation
- skip post-replacement verification
Next Step
After term decisions are stable, continue with /translate or /super-translate.
Example Usage
/term-decision
/term-decision Move
/term-decision "Basic Move"