Email, calendars, task management, and automation (7.9K skills)
Automate browser tasks such as clicking, typing, navigation, data extraction, testing, and screenshots. It controls Chrome through CDP, a browser control protocol, and supports work requiring logged-in sessions or pages rendered with JavaScript.
Use Cua Driver to operate desktop apps and browsers for testing, form filling, calendar entries, screenshots, and requested demo recordings. Inspect the target window before acting and check its state afterward to verify results.
Inspect and manage BB threads, environments, projects, and plugins through its command line interface. The skill guides agents to identify the correct targets, apply changes, and verify the resulting state.
Guide a setup process one step at a time in plain English. Each response provides the current actions and required values, followed by a short list of remaining stages that updates as the user progresses.
Coordinate a large job by splitting it into small tasks with explicit dependencies. Run ready tasks in parallel through Hermes, verify the combined result separately, and repair failures within defined limits before reporting completed work and remaining gaps.
Identify the MCP servers, command-line tools, API credentials, and connectors a workflow needs before work starts. The skill lists what is present, missing, optional, or unsafe and identifies the next setup or verification step.
Assess one project or product fact for future sessions. Prepare a memory candidate with its scope, tags, and retention, then decide whether to remember, refuse, or defer it without claiming an unverified memory update.
Create a metadata-only dashboard for an OMH skill collection. Bring together recorded changes, observed failure signals, pending amendments, and available updates, then identify responsible workflows and next actions without modifying the skills or treating usage counts as health.
Compare existing skills before creating a new one. This workflow organizes local and public candidates, source trust labels, and observed risk findings into a report that helps you choose between reuse, extension, and new development.
Choose an appropriate oh-my-hermes workflow for a request in Hermes Agent. It guides planning, research, and coding handoffs, explains the next action, and distinguishes prepared work from execution results that have actually been observed.
List the tools, command-line programs, API credentials, and service connections a workflow needs. Classify them as present, missing, optional, or unsafe, then identify the next verification step before work begins.
Turn short, dictated or mobile-style project requests into concise clarifications, plans, status updates, or handoffs. Resolve ambiguous intent and identify the next action, with confirmation required for risky steps and execution status kept explicit.
Resume previously agreed work from conversation context and track progress with a visible checklist. Keep one step active, preserve rejected ideas, distinguish blockers from user deferrals, and report completion only within the available evidence.
Manage long-term project goals through repeated interviews, planning, research, building, and review. Define bounded tasks, verify observed results, track constraints, and preserve resumable checkpoints when permissions, budgets, or external dependencies stop progress.
Manage local Hermes skills by listing and reading existing entries, creating or editing personal skills, and deleting them when needed. Get the appropriate commands for searching, installing, uninstalling, or hiding skills through Hermes.
Organize lessons from agent workflows into review records for memory, skills, wikis, failure retrospectives, or automation. Track metadata, prepare improvement proposals and regression cases, and audit skill writes against recorded reviews.
Structure live lookups for weather, prices, exchange rates, sports scores, local time, and directions. The skill specifies sources, timestamps, units, and freshness requirements, and distinguishes lookup preparation from results actually returned by a provider.
Capture lasting project facts, decisions, and preferences as small OMH-local memory records. Distinguish information ready to save from material that should be refused or reviewed, and report the observed storage status.
Ask Jev to recommend an OMH workflow when the route remains undecided. Present its top choice as a clarification question, record the response with its request identifier, and report the tool status and cost source.
Review existing Hermes memory for outdated, incorrect, duplicate, or conflicting claims. Confirm entries individually and prepare a concise change proposal, with explicit approval and Hermes's native memory tool required to apply changes.
Turn recurring corrections and workflow lessons into individual rule candidates with evidence, confidence scores, and defined scope. Review which belong within a project, which may apply across projects, and what is suitable for export.
Set up mail and calendar connections in Hermes for an on-demand morning summary. The workflow checks existing settings, guides authentication, and requires approval before applying changes, keeping access limited to reading and drafting.
Prepare a structured delivery policy for bot replies across Discord, Slack, and Telegram. Clarify the destination, thread, silent updates, and attachments, with explicit next steps and a clear distinction between planned delivery and confirmed sending.
Turn a task beginning with /omh into a recommended workflow or sequence of workflows. Consult the current OMH catalog, explain the selection, and identify the next action while keeping recommendations distinct from actual execution.