Email, calendars, task management, and automation (7.9K skills)
Review observations and unresolved conflicts from a knowledge-management workflow. Compare documented practices with actual settings, identify recurring problems, and propose specific changes with supporting evidence, risks, and reversibility details for human approval.
Adds a new subject area to an existing knowledge system. Through conversation, defines suitable terms, note fields, and processing methods, then creates folders and templates that connect with the existing structure after approval.
Helps an agent handle email, files, meetings, documents, and spreadsheets across 12 Google Workspace services. It selects the appropriate tools and guides workflows such as replying to messages, sharing files, and checking spreadsheet updates.
Refine a task before starting by checking the user's underlying goal, expert methods, and whether instructions are too loose or restrictive. Preserve decisions for the user, then carry out the work with an independent check of the result.
Ask the agent to restate its understanding of your goals and the problem you are trying to solve. The resulting explanation gives you a way to check how it has interpreted your request.
Summarize chat progress since the user's last message in plain language and small visuals. Gather every pending user decision with a recommendation, supporting evidence, and risks, so the user can decide directly from the reply.
Manages Codex Taskboard issues through taskctl, its command-line tool. It guides issue creation, comments, work claims, and status changes while checking conversation ownership and update conflicts before moving work forward.
Turn commits authored by the current Git user during a chosen period into a concise progress update. Highlight substantial behavior and architecture changes, include the actual dates, and leave out merges, uncommitted work, and cosmetic edits.
Restate the agent’s previous reply in simple, everyday language without jargon. Use it when an explanation feels difficult to follow and you want the same message expressed more clearly, briefly, and naturally.
Turn recurring work habits and communication preferences into a personal agent skill. It combines evidence from workspace conversations with direct questions, then drafts or updates instructions for how agents should work with you.
Keep reversible work moving while the user supervises asynchronously. Agents make reasonable execution decisions, complete edits, and present results for review, while still seeking confirmation for irreversible actions such as deleting production data or sending external messages.
Manage the limited conversation space an AI uses while working with long files or large outputs. Delegate bulk material to subagents, retain summaries in the main conversation, and limit each phase's scope and turn budget.
Rebuild recent working context from chat history, shared project records, and current branches or pull requests. Produce a concise brief showing progress, recurring problems, and one concrete next action before work resumes.
Keep a decision log for long or unattended tasks in a tab-separated file. Record choices, reasons, evidence, and outcomes, then check the log against the conversation and obtain a separate model’s review.
Run multiple agents in parallel, either on separate parts of a task or on competing attempts at the same brief. Combine their evidence into one report that identifies issues, missing coverage, and any workers that dropped out.
Send bug reports, suggestions, or other feedback to the Nexu team using `/feedback`. The submission script automatically includes conversation history and images, then reports whether the feedback was successfully forwarded.
Read Feishu chat history and replies within discussion threads, search across conversations by keyword, sender, or time, and download attached images and files. Retrieve additional pages when a complete conversation is needed.
Manage Feishu calendar events by creating or changing meetings, inviting participants, and checking team availability. You can also search events, respond to invitations, and check whether a meeting room reservation has been accepted.
Create and manage Feishu Bitable tables for work such as customer tracking. Add, search, update, or delete records, configure columns using 27 field types, filter and sort data, and manage views or batch imports.
Manage tasks and task lists in Feishu, including assignments, deadlines, and collaborators. Create to-dos, check unfinished work, mark tasks complete or reopen them, and organize recurring tasks within shared lists using the required plugin.
Start a probing interview to sharpen a plan or design. The skill invokes a questioning session through `/grilling`, giving you a way to work through an idea by responding to persistent questions from the agent.
Turns recurring work into bounded agent loops: repeated actions guided by fresh results and clear stopping rules. It can find existing loops, review or adapt them, run them with evidence, analyze completed runs, and save reusable prompts.
Turn recurring engineering work into feedback loops: repeatable steps with evidence checks and clear stopping rules. Discover opportunities, adapt published loops, repair existing designs, run authorized tasks, and review results for reuse.
Assemble a temporary team of two to five existing AI agents for a task. The skill selects relevant roles, assigns work, coordinates collaboration, and combines their outputs into a shared conclusion or proposal.