{"id":41585461,"url":"https://github.com/shiv19/endurance-coach-skill","last_synced_at":"2026-02-13T11:45:19.110Z","repository":{"id":334312379,"uuid":"1140922288","full_name":"shiv19/endurance-coach-skill","owner":"shiv19","description":"Endurance Coaching Skill for all AI agents","archived":false,"fork":false,"pushed_at":"2026-01-24T12:36:37.000Z","size":3264,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-01-24T12:58:33.753Z","etag":null,"topics":["coaching","endurance","running","swimming","training","triathlon"],"latest_commit_sha":null,"homepage":"https://shiv19.com/endurance-coach-skill/","language":"TypeScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/shiv19.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE.md","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":"ROADMAP.md","authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-01-23T23:50:24.000Z","updated_at":"2026-01-24T12:36:41.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/shiv19/endurance-coach-skill","commit_stats":null,"previous_names":["shiv19/endurance-coach-skill"],"tags_count":6,"template":false,"template_full_name":null,"purl":"pkg:github/shiv19/endurance-coach-skill","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/shiv19%2Fendurance-coach-skill","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/shiv19%2Fendurance-coach-skill/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/shiv19%2Fendurance-coach-skill/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/shiv19%2Fendurance-coach-skill/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/shiv19","download_url":"https://codeload.github.com/shiv19/endurance-coach-skill/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/shiv19%2Fendurance-coach-skill/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":28813229,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-01-27T12:25:15.069Z","status":"ssl_error","status_checked_at":"2026-01-27T12:25:05.297Z","response_time":168,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"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":["coaching","endurance","running","swimming","training","triathlon"],"created_at":"2026-01-24T09:03:03.698Z","updated_at":"2026-01-27T13:03:35.524Z","avatar_url":"https://github.com/shiv19.png","language":"TypeScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Endurance Coach\n\nEndurance Coach allows you to use Claude (or any AI assistant) to create custom-tailored training programs for triathlons, marathons, and other endurance activities. Using a data-driven approach and principles from top training plans, the AI will create a training plan that's uniquely fit for you, your personal fitness, and the constraints you have in the next couple of weeks. Maybe you're recovering from an injury, maybe you're traveling and don't have access to a pool or track in a certain week - tell the AI about it and it'll create a plan that works for you.\n\nThe output is a beautiful training plan app that allows you to add, edit, or move workouts, mark them as complete, and update key training data like heart rate zones, LTHR, threshold paces, FTP, and others. Your data is kept locally in your browser.\n\nWorkouts can be exported as simple calendar events (.ics), Zwift (.zwo), Garmin (.fit), or TrainerRoad/ERG (.mrc) workouts.\n\n## Examples\n\nSee example training plans at [shiv19.github.io/endurance-coach-skill](https://shiv19.github.io/endurance-coach-skill/#demos).\n\n## Installation \u0026 Creating a training plan\n\nThis skill works with any AI assistant that supports Skills (Claude.ai, Claude Code, and others). To use this tool, you need access to an AI assistant with network access for Skills. Depending on user/admin settings, Skills may have full, partial, or no network access.\n\nSyncing all your Strava activities and creating a tailored training plan takes ca. 15 minutes.\n\n### Installing the Skill\n\nInstall the skill using the following command:\n\n```bash\nnpx skills add shiv19/endurance-coach-skill\n```\n\nThis command works with any AI assistant that supports skills (Claude.ai, Claude Code, and others).\n\n**Additional Setup for Claude.ai:**\n\nIf you're using Claude.ai, you'll also need to:\n\n1. Open [Claude.ai Settings](https://claude.ai/settings/capabilities)\n2. Enable \"Code execution and file creation\"\n3. In the allowed domains list, add `*.strava.com`\n\n### Creating a plan\n\nUse the most capable model available to you. Prompt your AI assistant with something like this:\n\n\u003e Help me create a training plan for the Ironman 70.3 Oceanside on March 29th 2026 using the \"endurance-coach\" skill.\n\nYour AI assistant will ask how you'd like to provide your fitness data. You have two options: You can either tell the AI about your fitness history manually - or you can give it access to your Strava activities. I recommend the later - data doesn't lie and more data allows the AI to make a training plan that really fits you.\n\n#### Option 1: Connect to Strava (Recommended)\n\nThe easiest way to get a personalized plan is to let your AI assistant analyze your Strava training history. This gives the AI real data about your current fitness, training patterns, and progress.\n\nThe AI needs a `Client ID` and `Client Secret` to access your Strava activities. You're only giving the AI access to your data - nobody else gets to see it.\n\n1. Go to [strava.com/settings/api](https://www.strava.com/settings/api) and log in with your Strava account\n2. You'll see a form titled \"My API Application\" - fill it out:\n   - **Application Name**: Enter anything you like (e.g., \"Endurance Coach\")\n   - **Category**: Select \"Data Importer\"\n   - **Club**: Leave this blank\n   - **Website**: Enter any URL (e.g., `https://claude.ai`)\n   - **Application Description**: Enter anything (e.g., \"Training plan generation\")\n   - **Authorization Callback Domain**: Enter `localhost`\n3. Check the box to agree to Strava's API Agreement and click **Create**\n4. Copy your **Client ID** and **Client Secret** and give them to the AI when prompted\n\n#### Option 2: Manual Entry\n\nDon't use Strava, or prefer not to connect it? No problem. You can tell the AI about your fitness directly. Be prepared to share:\n\n**Current Training (recent 4-8 weeks):**\n\n- Weekly training hours by sport (swim/bike/run)\n- Typical long session distances (longest ride, longest run, etc.)\n- Training consistency (how many weeks have you been training regularly?)\n\n**Performance Benchmarks (any you know):**\n\n- Bike FTP (Functional Threshold Power) in watts\n- Run threshold pace or recent race times (5K, 10K, half marathon, etc.)\n- Swim CSS (Critical Swim Speed) or recent time trial (e.g., 1000m time)\n- Max heart rate and/or lactate threshold heart rate\n\n### Telling the AI about your event \u0026 constraints\n\nIn the next step, the AI will ask you about yourself, the event you're training for, and any constraints it should keep in mind. Examples of information you'd tell any coach:\n\n- Years in the sport\n- Previous races completed (distances and approximate times)\n- Any recent breaks from training\n- Injuries or health issues\n- Schedule limitations (work travel, family, etc.)\n- Equipment access (pool availability, trainer, etc.)\n\nThe AI will use this information to create a plan tailored to your current fitness level. The more detail you provide, the better your plan will be.\n\n## Contributing\n\nWe welcome contributions from the community! Whether you want to add workout templates, improve the UI, fix bugs, or enhance documentation, your help is appreciated.\n\nPlease see our [CONTRIBUTING.md](CONTRIBUTING.md) for:\n\n- Development setup instructions\n- Coding standards and guidelines\n- How to submit pull requests\n- Areas where we need help\n\n# About\n\n## Lineage \u0026 Architectural Evolution\n\nThis project originated as a fork of [Claude Coach](https://github.com/felixrieseberg/claude-coach) by [Felix Rieseberg](https://felixrieseberg.com), but it is no longer a fork in any meaningful architectural or behavioral sense.\n\nThe original project relied on large language models generating full, deeply nested workout plans as verbose JSON based solely on plain-text instructions. This approach had no formal schema contract, no validation loop, and no way for an AI agent to detect or correct structural errors before downstream rendering. Failures were late, brittle, and required human intervention.\n\nThis project deliberately replaces that architecture.\n\n### Key changes that make this an independent system:\n\n**Contract-first design**\nThe system now exposes explicit machine-consumable schemas (via validation tooling) rather than relying on prose instructions. AI agents can validate outputs and receive structured error feedback before proceeding.\n\n**Representation shift**\nWorkout plans are no longer generated as large raw JSON objects. Instead, the system defines a constrained domain-specific language composed of reusable workout templates. Plans are authored as concise YAML compositions of these templates, drastically reducing output size, entropy, and failure modes.\n\n**Template-based composition**\nThe AI agent does not invent workout structure freely. It selects from a predefined, inspectable set of workout building blocks supplied by the tool, turning generation into constrained composition rather than unconstrained construction.\n\n**Agent-first execution model**\nThe primary consumer is an AI agent, not a human. All commands are deterministic, side-effect explicit, and validation-first. Outputs are designed to support self-correction loops by the agent.\n\n**Independent surface and identity**\nThe project has a new name, a new installation path, and a different public contract. It is no longer Claude-specific and is designed to be consumed by any AI agent or orchestration framework.\n\nBecause the core abstraction, data representation, validation model, target consumer, and public interface have all changed, this project should be treated as an independently evolved system that acknowledges its lineage but does not share the original architecture or assumptions.\n\n## Disclaimer\n\nEndurance Coach is an independent, open-source project. It is not made by, endorsed by, or affiliated with Anthropic, PBC. \"Claude\" is a trademark of Anthropic. This skill works with Claude and other AI assistants but is developed and maintained independently.\n\n## Maintainer\n\nMaintained by [Shiva Prasad](https://shiv19.com) · [@multishiv19](https://twitter.com/multishiv19)\n\n## License\n\nMIT License. Original work Copyright © 2025 Felix Rieseberg. Modifications and independent development Copyright © 2025-2026 Shiva Prasad.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fshiv19%2Fendurance-coach-skill","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fshiv19%2Fendurance-coach-skill","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fshiv19%2Fendurance-coach-skill/lists"}