Email, calendars, task management, and automation (7.8K skills)
Automate website tasks with agent-browser, a command-line browser tool for AI agents. Fill forms, click controls, extract page text, capture screenshots, and test web apps, with support for saving and reusing login state.
Find installable agent skills for a specific task using the Skills CLI, a tool for managing skill packages. It explains relevant options, provides installation commands and information links, and can install a chosen skill when requested.
Explore websites with shell-style commands, treating URLs as paths and page elements as files. List links, search text, extract selected content, and navigate cached pages, with background fetching and configurable link prefetching.
Connect SharpAI Aegis's Clawdbot agent to a Matrix/Element room to search clips, receive alerts, and control cameras through conversation, with support for files, reactions, message edits, and threaded replies.
Send Aegis activity detections to Home Assistant as events that trigger automations. Use camera-specific conditions to connect detected activity with actions such as turning on lights or sending notifications.
Send SharpAI Aegis events as JSON to a webhook, a URL that receives event notifications. Choose which events to send and connect camera activity to Zapier, IFTTT, Make, n8n, or a custom endpoint.
Publish SharpAI Aegis camera events through MQTT, a messaging protocol used by automation platforms. Choose event types and connection settings to feed clip analysis, person detection, alerts, or camera outages into compatible workflows.
Connect the Clawdbot agent in SharpAI Aegis to LINE Messenger to send messages and files, and reply to incoming messages through a messaging channel configured with LINE credentials.
Connect the Clawdbot agent in SharpAI Aegis to Signal using signal-cli. The channel supports sending messages and files, replying, and adding reactions, with settings for the command path and bot phone number.
Use a browser to research topics, extract page information, fill forms, and test websites. Guides agents through identifying page controls, interacting with web apps, saving screenshots or PDFs, and reusing saved login sessions.
Build flexible daily plans that account for ADHD-related difficulties with time estimation and getting started. Narrow priorities to three tasks, add transition buffers, and choose small next actions that match the energy available that day.
Automate laboratory data management in Benchling through its Python SDK and REST API. Manage biological sequences, sample inventory, electronic lab notebooks, and workflow tasks, with guidance for external data synchronization and SQL reporting.
Keep preferences, procedures, and performance findings in knowledge pages that persist across sessions. Hindsight updates pages from retained conversations, while the available tools let agents retrieve pages, search memories, and ingest documents or local files.
Organize project memory so agents can retrieve context across sessions. Review priorities and source records, consolidate duplicate notes, archive stale material, or carry out authorized erasure while preserving the authority of official records and documenting affected data.
Find skills on SkillsMP or GitHub, install them for one agent or globally, and inspect available skills. Create custom skills with SKILL.md metadata, register them in agent.yaml, and validate the agent configuration.
Answer questions using an existing wiki, combining relevant pages with citations to both wiki entries and original documents. Identify gaps and confidence levels, and optionally save useful answers as new pages after user approval.
Organize tasks into small work packets, review them, and dispatch approved work with execution logs. Tasks that change external state require approval, and added scope is handled through a new packet.
Keep goals, decisions, evidence, and next actions organized during long or complex agent tasks. Use persistent records and checkpoints to resume work, coordinate agents, and verify changes across files before delivery.
Keep parallel Claude Code sessions identifiable with project, branch, and task labels. Report clear progress states at major milestones, explain blockers, and include a suggested default when asking for missing information.
Find the right pro-workflow skill or command for a task. This index groups tools by job, explains when to use each one, and distinguishes skills you invoke yourself from those the agent selects automatically.
Create a handoff document that records completed work, unfinished tasks, key decisions, changed files, and potential pitfalls. Include a ready-to-paste instruction so the next agent session can resume with the context it needs.
Review saved lessons and correction history before starting a task. The skill ranks relevant findings, explains earlier mistakes, highlights sessions with frequent corrections, and suggests an approach based on those records.
Review connected MCP servers, which give agents access to external tools, for unused or overlapping capabilities. Get recommendations to keep, disable, or review each server, with estimates of the request size each recommendation could save.
Turn mistakes and feedback into specific rules for future sessions. The skill records the category, mistake, and correction, proposes an actionable rule, and saves it to project memory only after approval, with optional scope for a single wiki.