How this skill is triggered — by the user, by Claude, or both
Slash command
/novadb-consulting:create-commentThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
**This skill ONLY handles:** Creating a new comment on an object in NovaDB.
This skill ONLY handles: Creating a new comment on an object in NovaDB.
For updating existing comments → use update-comment
Create a new comment on an object in NovaDB.
Note: NovaDB object IDs start at 2²¹ (2,097,152). All IDs in examples below are samples — always use real IDs from your system.
novadb_cms_create_comment — Create the commentnovadb_cms_get_comment — Fetch the created comment (required follow-up){
"branchId": 2100347,
"objectRef": 12345,
"body": "<div>My comment text</div>",
"username": "jdoe"
}
branchId — Branch ID (number, required)objectRef — Object ID to comment on (number, required)body — Comment body as XHTML (string, required, see body rules below)username — (optional) Acting username for auditThe comment body must be valid XHTML with a <div> root element.
<div>user text here</div><div> or <div , use it as-is"<div>Great work!</div>"<div> if needed)novadb_cms_create_comment with the XHTML body{ id } — not the full commentnovadb_cms_get_comment with the returned id to fetch the full commentThe create call returns { id }. After fetching, the full comment includes id, body, branch, object reference, author, timestamps, etc.
Comment body must be valid XHTML with a <div> root element. Wrap plain text as <div>user text</div>.
Returns { id }. Use the ID with get-comment to fetch the full comment data.
npx claudepluginhub novadb/claude-plugins --plugin novadb-consultingPosts comments to Xiaohongshu feeds via xiaohongshu-mcp HTTP API using Node.js script. Requires feed_id, xsec_token, and content via environment variables. Useful for social media automation.
Adds, lists, edits, or deletes comments on Jira issues via explicit /comment-jira-issue invocation. Requires user-provided body content and confirmation for write actions.
Searches MemPalace before answering questions about past work, people, projects, or prior decisions. Returns verbatim stored content instead of guessing from model memory.