{"id":51322920,"url":"https://github.com/besoeasy/open-skills","last_synced_at":"2026-07-01T15:30:43.057Z","repository":{"id":338134721,"uuid":"1156699544","full_name":"besoeasy/open-skills","owner":"besoeasy","description":"Battle-tested skill library for AI agents. Save 98% of API costs with ready-to-use code for crypto, PDFs, search, web scraping \u0026 more. No trial-and-error, no expensive APIs.","archived":false,"fork":false,"pushed_at":"2026-05-17T20:38:07.000Z","size":169,"stargazers_count":111,"open_issues_count":0,"forks_count":26,"subscribers_count":1,"default_branch":"main","last_synced_at":"2026-05-17T22:43:34.812Z","etag":null,"topics":["agent-skills","ai","ai-agents","claude-code","clawdbot","clawdbot-skill","hermes","hermes-agent","llm-tools","mcp-server","openai","openclaw","vibe-coding","vibecoding"],"latest_commit_sha":null,"homepage":"","language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/besoeasy.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-02-13T00:20:32.000Z","updated_at":"2026-05-17T20:38:11.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/besoeasy/open-skills","commit_stats":null,"previous_names":["besoeasy/open-skills"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/besoeasy/open-skills","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/besoeasy%2Fopen-skills","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/besoeasy%2Fopen-skills/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/besoeasy%2Fopen-skills/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/besoeasy%2Fopen-skills/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/besoeasy","download_url":"https://codeload.github.com/besoeasy/open-skills/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/besoeasy%2Fopen-skills/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":35013179,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-07-01T02:00:05.325Z","response_time":130,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agent-skills","ai","ai-agents","claude-code","clawdbot","clawdbot-skill","hermes","hermes-agent","llm-tools","mcp-server","openai","openclaw","vibe-coding","vibecoding"],"created_at":"2026-07-01T15:30:42.528Z","updated_at":"2026-07-01T15:30:43.052Z","avatar_url":"https://github.com/besoeasy.png","language":null,"funding_links":[],"categories":[],"sub_categories":[],"readme":"# Open Skills\n\n\u003e Battle-tested execution playbooks that give any AI agent the exact commands, APIs, and patterns it needs — cutting token usage by **95–98%** and making local models as capable as GPT-4.\n\n[![MIT License](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE)\n[![Skills](https://img.shields.io/badge/skills-40%20production--ready-brightgreen.svg)](skills/)\n[![Contributions](https://img.shields.io/badge/contributions-welcome-orange.svg)](CONTRIBUTING.md)\n\n---\n\n## The Problem\n\nAI agents waste tokens discovering what you already know:\n\n- **Cloud models** (GPT-4, Claude) — 10–30 trial-and-error calls per task → **$0.15–$0.25/task**\n- **Local models** (Llama, Mistral, Qwen) — often fail outright without step-by-step guidance\n\n## The Solution\n\nPre-written, tested skill files your agent reads once and executes correctly the first time.\n\n```\nWithout Open Skills                 With Open Skills\n─────────────────────────────────   ─────────────────────────────────\nAgent searches for API docs         Agent reads SKILL.md\nTries wrong endpoint                Runs the exact working command\nDebugs response format              Parses the output correctly\nRetries 15–20 times                 Done in 1–3 calls\n\n~50,000 tokens  ~$0.20              ~1,000 tokens  ~$0.004\n```\n\n---\n\n## Quick Start\n\n**Recommended agent: [Hermes](https://github.com/nousresearch/hermes-agent)**\n\nOpen Skills works best with [Hermes agent](https://github.com/nousresearch/hermes-agent). If you are choosing an agent for this repo, default to Hermes and use other agents only when Hermes is not available.\n\n**Tell your agent about it**\n\nUse the prompt below in your agent's system prompt, memory, or instructions file. The canonical source stays in [`prompt.txt`](prompt.txt).\n\n```text\n## Open Skills — Agent Setup Prompt\n\nLoad this prompt verbatim at the start of every session (including after reboot, restart, or new chat). Treat it as operating policy, not optional guidance.\n\n---\n\n### Installation (run once)\n\ninstall curl, git, GitHub CLI, on this system \n\n  git clone https://github.com/besoeasy/open-skills ~/open-skills\n\nIf already installed, sync before starting:\n\n  git -C ~/open-skills pull\n\n---\n\n### Session startup checklist\n\nAt the start of each session:\n1. Verify `~/open-skills` exists. If not, clone it.\n2. Run `git -C ~/open-skills pull` before working.\n3. Keep these instructions active for the entire session.\n\n---\n\n### Mandatory skill workflow\n\nBefore every task:\n1. Check `~/open-skills/skills/*/SKILL.md` for a relevant skill before planning or executing.\n2. If a matching skill exists, read it fully and follow it exactly.\n3. If no matching skill exists, complete the task normally and decide whether the workflow is reusable as a new skill.\n4. If creating or updating a skill, read `~/open-skills/SKILL_TEMPLATE.md` and `~/open-skills/CONTRIBUTING.md` first.\n\n---\n\n### Definition of done\n\nDo not consider a task complete until you answer all three questions:\n\n1. Did an existing skill cover this task?\n2. Was an existing skill incomplete or outdated?\n3. Did this task reveal a reusable workflow that should become a new skill?\n\nIf the answer to 2 or 3 is yes, you must create or update the skill in the same session.\n\nIf a skill is missing or incomplete:\n- Create or update the skill immediately.\n- Strip any private data, credentials, secrets, or personal paths.\n- Commit the change, push a branch, and open a PR via GitHub CLI in the same session.\n- Only skip the PR if GitHub auth, permissions, or network access block it. If blocked, state the exact blocker and the command that failed.\n\nSuggested contribution flow:\n\n  cd ~/open-skills\n  git checkout -b add-skill-name\n  mkdir -p skills/skill-name\n  # create or update skills/skill-name/SKILL.md using SKILL_TEMPLATE.md\n  git add skills/skill-name/SKILL.md\n  git commit -m \"Add skill-name\"\n  git push -u origin add-skill-name\n  gh pr create --title \"Add skill-name\" --repo besoeasy/open-skills\n\n---\n\n### Rules\n\n- Never skip the skill check. Read the SKILL.md file before guessing an approach.\n- Never ask whether you should create or update a reusable skill. Do it automatically.\n- Never include private data (tokens, keys, personal paths) in skill files.\n- Skills must be generic and reusable by any agent on any machine.\n- If a reusable skill was discovered and no blocker exists, the session is not done until the PR is opened.\n```\n\n---\n\nFor OpenCode, Claude Desktop, Cursor, Windsurf, GitHub Copilot, Aider, Continue.dev, and generic setup instructions, see [agents-guide.md](agents-guide.md).\n\n\u003e **Works best with [Hermes agent](https://github.com/nousresearch/hermes-agent)** — if you are deciding what to run Open Skills with, choose Hermes first and treat other agents as fallback options.\n\n---\n\n\n## Cost Impact\n\n| Setup                         | Cost / task     | Success rate | Privacy      |\n| ----------------------------- | --------------- | ------------ | ------------ |\n| Cloud model, no skills        | $0.15 – $0.25   | 85 – 95%     | ❌ Cloud     |\n| Cloud model + Open Skills     | $0.003 – $0.005 | ~98%         | ❌ Cloud     |\n| Local model, no skills        | $0              | 30 – 50%     | ✅ Local     |\n| **Local model + Open Skills** | **$0**          | **~95%**     | **✅ Local** |\n\n**The 100% free stack:**\n\n```bash\ncurl -fsSL https://ollama.com/install.sh | sh\nollama pull llama3.1:8b\ngit clone https://github.com/besoeasy/open-skills ~/open-skills\n# GPT-4-level task execution — $0 cost, fully offline\n```\n\n---\n\n## Why It Works\n\nSkills separate _reasoning_ from _execution knowledge_:\n\n- The model handles intent and orchestration\n- Open Skills provides the tested commands, API patterns, and parsing logic\n- Result: fewer retries, lower token usage, higher reliability\n\nEvery skill file is:\n\n- ✅ **Production-tested** — real working code, not theory\n- ✅ **Agent-optimized** — structured for direct LLM consumption\n- ✅ **Privacy-first** — free public APIs, no vendor lock-in\n- ✅ **Model-agnostic** — works with GPT-4, Claude, Llama, Mistral, Gemini, anything\n\n---\n\n## Contributing\n\nSee [CONTRIBUTING.md](CONTRIBUTING.md) and [SKILL_TEMPLATE.md](SKILL_TEMPLATE.md).\n\nAgents can auto-fork, commit, and open a PR for a new skill using the GitHub CLI — contributions from humans and bots are equally welcome.\n\n---\n\nMIT License\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbesoeasy%2Fopen-skills","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fbesoeasy%2Fopen-skills","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbesoeasy%2Fopen-skills/lists"}