Skip to content
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 5 additions & 2 deletions .github/agents/Draft-doc.agent.md
Original file line number Diff line number Diff line change
Expand Up @@ -39,9 +39,12 @@ For each documentation request (update, addition, or creation):
- Load the product-specific variable set from `Project/VariableSets/{Product}.flvar` (e.g., `Replicate.flvar`, `Sense_Release.flvar`) and use variables for product names, versions, and service names (see copilot-instructions.md for details).
- Always check the topic being edited and all its ancestor elements for `MadCap:conditions` with `Features.<ticket-id>` or `Features.NotIn<ticket-id>` before editing. If found, invoke the **flare-feature-conditions** skill and follow its decision table.
- Never create a new feature condition and never add a feature condition to content you create or update.
- **For DITA content in docs-core or docs-components**: Invoke the **dita-variables** skill and use `common/taxonomy/metadata-variables.dita` `<keyword conref>` references for product names, module names, and brand names.
- **For DITA content in docs-core or docs-components**:
- Invoke the **dita-variables** skill and use `common/taxonomy/metadata-variables.dita` `<keyword conref>` references for product names, module names, and brand names.
- Treat prompt, handoff, Jira, and SME wording as factual input, not markup-ready text. Do not carry hardcoded product, module, service, engine, or brand names from those sources into DITA. Resolve each name by invoking the **dita-variables** skill before drafting.
- Before delivery, scan every new or edited DITA topic for hardcoded names that require variables. XML parsing alone is not sufficient validation.
- Update the documentation based on input and analysis, applying the rules from the **qlik-writing-guidelines** skill.
- Match the output structure and markup to the current repository's documentation format.
- Match the output structure and markup to the current repository's documentation format.
- For unresolved links, cross-references, and images, follow the `[ASSUMED-*]` pattern documented in copilot-instructions.md:
- Use empty `href=""` or `src=""` attributes (never dummy/placeholder values that will break the build)
- Place `[ASSUMED-LINK]`, `[ASSUMED-XREF]`, or `[ASSUMED-IMAGE]` labels visibly in the content text
Expand Down
Loading