Find your next open-source project

Search by name, purpose, or keyword — or paste a GitHub repository URL
All skills

rlaope/oh-my-hermes

Adds a large set of ready-made skills to the Hermes coding harness so an agent can plan, build, review, research and operate software work inside your IDE.

  • 127 skills
  • 4.7k stars

The 5 most installed

Ordered by installs on skills.sh
  • review-sweep38 installs

    Reviews open oh-my-hermes pull requests under REVIEW.md and posts findings as GitHub reviews.

  • triage-sweep38 installs

    Backfills labels across unlabeled oh-my-hermes issues and pull requests.

  • Classifies build, typecheck and CI failures and points to the likely cause.

  • Analyzes a dataset or table you supply and reports what it shows.

  • Finds swallowed or hidden errors in code that may be masking failures.

The other 122 skills
  • omh-ops-review35 installs

    Reviews recurring operating status, risks and blockers and reports where things stand.

  • Turns product evidence into a PRD or roadmap priority decision.

  • ulw-perf35 installs

    Investigates software slowness, memory leaks and cost spikes.

  • Checks a web page for screen-reader and keyboard accessibility gaps against WCAG.

  • Cleans up messy or AI-generated code by removing filler and tightening what remains.

  • omh-backend34 installs

    Helps design APIs, servers and data-layer changes.

  • Refreshes an outdated local code index or codemap.

  • omh-decide34 installs

    Compares options for a team or business decision and recommends one.

  • Evaluates a plugin, connector or API before you adopt it.

  • omh-files34 installs

    Cleans up folders and handles renames or deletions under a set policy.

  • omh-frontend34 installs

    Builds or polishes a web or terminal user interface.

  • Handles incoming GitHub pull requests, issues and CI events.

  • omh-image-cards34 installs

    Creates an infographic card summarising meeting notes, a report or a pull request.

  • omh-llm-app-dev34 installs

    Helps build an LLM-powered feature, from prompt setup to integration.

  • omh-media-input34 installs

    Processes audio, video or screenshots you send in.

  • Explains a supplied research paper or PDF so you can understand its key ideas.

  • Breaks a decided cross-module refactor into ordered phases.

  • omh-rust34 installs

    Prepares fixes for Rust ownership, lifetime or unsafe code problems.

  • Reviews prompts, tools and automation for agent safety risks.

  • Audits accumulated technical debt and suggests where to deepen the architecture.

  • omh-terminal34 installs

    Runs shell commands and test suites under a policy overlay that constrains what is allowed.

  • Lists which tools and credentials a workflow needs before it can run.

  • ulw-loop34 installs

    Runs many build cycles toward an ambitious project goal.

  • ulw-plan34 installs

    Prepares a high-stakes technical proposal for your approval before work starts.

  • ulw-research34 installs

    Researches a topic in depth and gathers grounded sources before you make a decision.

  • model-onboarding33 installs

    Onboards a newly released AI model into the harness, including routing, pricing and benchmark checks.

  • omh-agent-debug33 installs

    Diagnoses an agent that is stuck, looping or drifting and captures the failure state.

  • Summarises AI agent progress so managers can inspect what agents did and where they stalled.

  • omh-apps33 installs

    Performs actions in external apps such as email, Slack or Jira.

  • omh-ask33 installs

    Consults an outside AI advisor for a second opinion on your work.

  • omh-cancel33 installs

    Ends an active workflow and clears its running state.

  • omh-code-review33 installs

    Reviews a pull request or change set with bugs as the first priority.

  • Gives a guided tour of an unfamiliar repository so you can start working in it.

  • Writes or rewrites product and company copy for publishing.

  • Plans a compact context when the token budget or context window is at risk.

  • omh-cto-loop33 installs

    Reviews engineering roadmaps and risks from a leadership perspective and suggests priorities.

  • Builds a course syllabus and training plan for a team or class.

  • Brings back previously rejected options and the reasons they were dropped.

  • Tracks the status of attachments and deliverables such as PPT and PDF files.

  • Guides a release rollout with a checklist and health signals to watch.

  • Checks a deliverable against a premium polish standard before it ships.

  • Clusters unsorted customer feedback and bug reports into themes.

  • Investigates budget overruns, cash risk or accounting close issues.

  • Lists and normalizes agent sessions and worktrees so you can find lost work.

  • Turns a rough app idea into concrete decisions and a path toward launch.

  • Collects repeated project lessons and promotes them into reusable guidance.

  • omh-jit-learn33 installs

    Teaches you just enough of a missing topic to unblock your current project.

  • omh-live-info33 installs

    Looks up live rates, scores or local times a work task needs.

  • Reviews translated product or content releases for localization quality before shipping.

  • Produces PowerPoint, PDF, Excel or HWP deliverables from your content.

  • Prepares an agenda and talking points for an upcoming meeting that lacks one.

  • omh-memory-new33 installs

    Saves a project fact so it is remembered in future sessions.

  • Keeps team meeting minutes, retrospectives and decision history organised.

  • Tracks cost, tokens, latency and service health in one view.

  • omh-people-ops33 installs

    Structures your company's hiring or HR process into clear steps.

  • Checks readiness before an agent controls a robot, 3D printer or IoT relay.

  • omh-plan33 installs

    Plans a software feature or bugfix that has not been planned yet.

  • Evaluates release readiness before an imminent production launch.

  • Reviews provider credentials and settings without exposing secrets.

  • Writes postmortems and reviews SLOs after an outage or missed target.

  • Builds a weekly or monthly report package for executives.

  • Researches a business question about markets, competitors or pricing and writes a brief.

  • Turns repeated lessons into written rules the agent can follow.

  • Reports on run efficiency and context usage for supplied sessions.

  • Shows a live board of which coding units are currently running.

  • Turns a prospect or market into a plan for pursuing the account.

  • omh-skill-health33 installs

    Gives a metadata-only health overview of your whole skill collection.

  • omh-skill-scout33 installs

    Finds existing skills you can reuse before writing a new one.

  • Gathers candidate papers, datasets or repositories for a research task.

  • Helps answer or escalate support tickets your team must handle.

  • omh-visual-qa33 installs

    Prepares observed-only evidence so rendered UI can get a visual verdict.

  • omh-voice-input33 installs

    Turns a dictated voice note about project work or status into usable input.

  • omh-wiki33 installs

    Builds a wiki from scattered knowledge so it is organized in one place.

  • Maps your repository, skills and setup to inventory the workspace and find gaps.

  • ulw-context33 installs

    Clarifies unclear or inconsistent vocabulary used across a repository.

  • ulw-interview33 installs

    Asks one question at a time to turn a vague request into a clear spec.

  • ulw-qa33 installs

    Runs hostile scenario tests and fix loops to harden software.

  • ulw-work33 installs

    Implements an accepted plan step by step until the work is done.

  • Puts a technical proposal through adversarial review to surface weak points before commitment.

  • Diagnoses native program crashes and memory corruption.

  • Scores a web page against design-award criteria and points out what to improve.

  • omh-codebase-uml31 installs

    Turns a repository into architecture diagrams.

  • omh-docs31 installs

    Explains how this harness works, based on the current source code.

  • Restructures an oversized UI component without changing its behavior.

  • Turns a chat report into a well-formed public GitHub issue.

  • Helps choose a serving engine and set up self-hosted LLM serving on GPUs.

  • omh-web-research31 installs

    Finds and cites current technical or business facts from the web.

  • omh-apple-design30 installs

    Designs or reviews iOS, macOS and other Apple platform interfaces.

  • Tests whether an early product idea has real customer demand before you build it.

  • Turns a running system into a map of possible attack paths.

  • Runs a small bounded spike to settle an uncertain technical choice.

  • Turns churn or weak activation signals into growth experiments for users.

  • Reviews a CRM pipeline or sales forecast and highlights what needs attention.

  • Helps respond to an open production incident or outage step by step.

  • Responds to a security event affecting code that is already shipped.

  • Traces wrong values, flaky tests and other misbehaviour in application code.

  • omh-iac-change20 installs

    Plans and reviews infrastructure-as-code changes in Terraform, OpenTofu or Pulumi before you apply them.

  • Helps with relational database work such as slow queries, indexes and schema changes.

  • Writes and maintains agent instruction files such as AGENTS.md and CLAUDE.md for a repository.

  • Writes commit messages and pull-request descriptions for your changes.

  • Helps build and fix ETL or streaming data pipeline jobs.

  • omh-git-workflow18 installs

    Rescues a troubled Git branch, merge conflict or broken commit.

  • Guides testing of internal controls such as SOX, ICFR or ITGC.

  • Walks through releasing an iOS or Android app to the app stores.

  • Guides fine-tuning a model on your own data with SFT, DPO or RLVR.

  • omh-release-cut18 installs

    Tags, stages and ships a versioned release behind a controlled rollout.

  • Checks a dataset, log or export for sensitive content before it leaves the team.

  • omh-spend-gate15 installs

    Pauses before spending your money on a checkout or paid API call and asks for approval.

  • Sizes a market and analyses customers and competitors.

  • omh-show-reel11 installs

    Makes a short video such as an explainer or motion piece from an idea or product.

  • Retrieves and reconciles public evidence behind biomedical variant claims.

  • Runs coordinate-safe polygon overlays and other spatial layer analysis.

  • Runs bounded standalone Vina docking on prepared PDBQT inputs.

  • omh-repro-audit10 installs

    Compares evaluation logs against the conditions reported in a paper.

  • omh-sprite-forge10 installs

    Turns image-model sprite sheets into animated character art for a game.

  • Screens literature against frozen inclusion criteria for a systematic review.

  • Helps shrink or cheapen a model by compressing its weights.

  • omh-skill-eval1 installs

    Checks a skill before you ship it by running the skill-eval command and reporting problems.

  • Breaks a whole product design problem into bounded tasks and delegates them.

  • Defines and records build and test proof that a change is done before merge.