useful-scripts

Read pages from Atlassian Confluence Cloud (CCAOA tenant by default). Use when the user asks you to review, summarize, fetch, audit, or inspect a Confluence page or its children — anything starting with a URL like https://ccaoa.atlassian.net/wiki/spaces/.../pages/... or a page ID. Handles auth via the shared atlassian_auth helper and prefers the v2 REST API. Confluence Cloud has no v3 — do not look for one.

Create or update pages in Atlassian Confluence Cloud (CCAOA tenant by default). Use when the user asks you to write, draft, publish, post, create, update, edit, or "fix" a Confluence page. Two paths: hand a markdown file to the existing md_to_confluence.py CLI for whole-page writes, or do surgical raw-HTTP edits in storage format for in-place table-row / status-macro / single-section changes.

Add a comment to an existing Jira Cloud issue (CCAOA tenant by default). Use whenever the user asks you to comment on, add a note to, leave an update on, reply on, or "post to" a Jira ticket — phrasings like "add notes to PROJ-123", "comment on that ticket", "drop a status update on the Jira", or "leave a note for the team on DEVOPS-1". Wraps jira_comment.py, which posts via the v2 (wiki-markup) or v3 (ADF) comment API using the shared atlassian_auth helper. Do not invoke for reading an issue (use jira-get), searching (use jira-read), creating a new ticket (use jira-create), or editing description/status fields.

Create a new Jira Cloud issue (CCAOA tenant by default). Use when the user asks you to file, open, log, create, draft, or "make a Jira" for a new ticket, bug, task, or epic. Wraps the existing jira_create_issue.py CLI in useful-scripts; for "epic + N tasks" bulk creation, points to jira_create_epic_with_tasks.py in the same directory. Handles auth via the shared atlassian_auth helper. Do not invoke for reading (use jira-get) or searching (use jira-read).

Fetch a single Jira Cloud issue by key or URL (CCAOA tenant by default). Use when the user asks you to read, look at, summarize, inspect, or pull the details of a Jira ticket — anything starting with `PROJ-123` or a URL like https://ccaoa.atlassian.net/browse/PROJ-123. Handles auth via the shared atlassian_auth helper and prefers Jira Cloud REST v3 (ADF). Do not invoke for Confluence pages (use confluence-read), JQL / multi- issue search (use jira-read), or creating a new ticket (use jira-create).

Search Jira Cloud issues via JQL (CCAOA tenant by default). Use when the user wants to list, find, query, or filter Jira tickets — phrasings like "all open tickets in X", "what's assigned to me", "tickets created this week", "everything tagged with Y", or a raw JQL string. Handles auth via the shared atlassian_auth helper and uses Jira Cloud REST v3. Do not invoke to fetch one specific ticket (use jira-get) or to create one (use jira-create).

Turn a raw dataset (CSV, Excel, JSON) into a decision brief with charts. Trigger this skill whenever the user has a data file they want analyzed, profiled, or summarized — including phrases like "explore this CSV", "profile this dataset", "what's in this spreadsheet", "build me a decision brief from this data", "analyze this Excel file", "find issues in this data", "summarize this dataset for the team", or "I have a dataset and need to make a decision". Also trigger on "data quality check", "data audit", or "give me an exec summary of this data". Even when the user doesn't explicitly say "brief" — if they have a tabular file and an unclear path forward, this skill is the answer.

Generate a structured content brief from a Substack Live (or similar live session / webinar / podcast / panel) transcript. Trigger this skill whenever the user mentions a Substack Live, a live session transcript, a recorded webinar, a panel discussion, or asks for a content brief / repurposing ideas / standalone quotes / "key themes" pulled from a recording — even if they don't say the word "Substack". Also trigger on phrasings like "I just finished a live and need to turn it into content", "pull quotes from this transcript", "summarize my session", or "what should I post from this".

Write a session handoff to .claude/session-handoff.md so the next Claude Code session can pick up where this one left off. File rotation is fully automated — never ask the user about it.