Hyperhuman Agent Workspace
Hyperhuman v1.0.0
Publisher description
From the marketplace listing
Work with Hyperhuman fitness content directly in ChatGPT. Explore sample workouts, review structured workout data, inspect exercise flows and media, and generate copy-ready API or embed snippets for web and mobile apps. Sign in with Hyperhuman Teams to browse your workspace content, manage workouts, programs, exercises, and social drops, and run selected content workflows through chat.
Language: English · Automatically detected from descriptions.
Files & skills
File archives
Skill instructions
adapt-for-member1.55 KB
--- name: adapt-for-member description: > Recommends or adapts Hyperhuman workouts and programs for a stated member profile. Use when the user asks to personalize, recommend the next session, adapt intensity, swap equipment, or reshape a plan for goals, level, or duration. Not for medical advice, clinical rehab plans, pricing, checkout, or social drops. --- # Adapt for a member ## Goal Produce a safer, science-informed next session or plan change from explicit user inputs. ## Steps 1. Collect only what the user provides: goal, level, duration, equipment, constraints, optional source workout/plan id. 2. If no source id: call `recommend_workouts` or `recommend_plans` with the stated filters. 3. If a workout id is known: call `adapt_workout` with profile fields the user gave. 4. If a program/plan id is known: call `adapt_plan`. 5. Optional: call `generate_workout` only when they ask for a new session (not a tweak). 6. Summarize: what changed, why (short), and one clear next step. ## Output - Recommended or adapted item name + id - 2–4 change bullets (duration, difficulty, equipment, substitutions) - Explicit non-medical disclaimer when health context appears ## Guardrails - Movement support only. Not medical care. Refuse diagnosis/treatment/prescription. - GLP-1 / weight management = adherence and lean-muscle movement support, not clinical guidance. - Do not invent end-user health records or pull insights digests. - Prefer conservative adaptations when constraints are unclear. - See [references/safety-guardrails.md](references/safety-guardrails.md).
Referenced files: 5
build-workout1.4 KB
--- name: build-workout description: > Builds a Hyperhuman team workout from library exercises. Use when the user asks to create a workout, assemble exercises and rests, or draft a session in their Hyperhuman team library. Requires OAuth. Not for guest sample browse, embeds, social drops, pricing, or medical advice. --- # Build a workout ## Goal Create a reviewable team workout from explicit structure choices. ## Steps 1. Confirm OAuth. If missing, stop and ask the user to connect Hyperhuman. 2. Gather: name, duration intent, category/difficulty, locale if needed. 3. Load options as needed: `list_team_goals`, `list_workout_categories`, `list_equipment`, `list_muscle_groups`, `list_team_exercises`. 4. For each step: `create_exercise_instance` or `create_rest_instance` with clear duration/reps. 5. Call `create_workout` with the collected `exercise_map` / instances and metadata. 6. Offer `update_workout` only if the user requests edits. 7. Reply with workout id, name, step count, and one next action (open in Teams or adapt). ## Output - Created workout id + name - Short structure summary - Link guidance: manage at team.hyperhuman.cc ## Guardrails - Explicit invocation preferred; do not invent team content when unauthenticated. - Respect team library only; do not claim stock was added unless a tool did it. - No pricing or checkout. - See [references/create-flow.md](references/create-flow.md).
Referenced files: 5
make-social-drop1.47 KB
--- name: make-social-drop description: > Creates and exports a Hyperhuman social drop from a workout, program, or exercise. Use when the user asks for Reels, Shorts, TikTok clips, captions, covers, or social drop export. Requires OAuth. Not for direct posting to social networks, pricing, checkout, embeds, or medical advice. --- # Make a social drop ## Goal Produce a channel-ready drop package (video variants, cover, copy) and export it. ## Steps 1. Confirm OAuth. If missing, stop and ask the user to connect Hyperhuman. 2. Identify source: workout, program, exercise, or exercise group id. Browse with `list_team_workouts`, `list_team_programs`, or `list_team_exercises` if needed. 3. Call `create_drop` with `source_type` and source id(s). 4. Wait for ready state; use `list_team_drops` / status fields from tool results. Do not claim Ready if the tool says Processing. 5. On request: `regenerate_drop_copy` and/or `regenerate_drop_cover`. 6. Call `export_drop` when the user wants downloadable assets. 7. Summarize: drop id, channel presets, export links/copy pack, quota note if returned. ## Output - Drop id + status - What was generated (ratios / durations / copy) - Clear note: export does not auto-post to Instagram/TikTok/YouTube/Facebook ## Guardrails - Explicit invocation preferred (writes + publish-adjacent). - `export_drop` records usage; it does not schedule Automate social publish. - No pricing or checkout. - See [references/drop-flow.md](references/drop-flow.md).
Referenced files: 5
ship-content-api1.75 KB
--- name: ship-content-api description: > Ships Hyperhuman workouts into a web or mobile product as a multi-step integration workflow (embed vs stream vs custom player). Use only when the user explicitly asks to ship or implement Hyperhuman in their own app as a full integration path. Not for sample workout browsing, standalone embed-code snippets, health checks, team library admin, creating workouts, social drops, pricing, checkout, or medical advice. --- # Ship in my app ## Goal Give a clear path from sample content to a copy-ready integration plan. ## Steps 1. If the user only asked for sample workouts, do not run this skill; that is `list_sample_workouts`. 2. If the user only asked for embed HTML, do not run this skill; that is `get_embed_code_samples`. 3. Clarify platform (`web` | `mobile` | `hardware`) and use case only when they asked to ship or integrate. 4. Call `get_integration_plan` with `platform` and `use_case`. 5. Call `list_sample_workouts` only if they also asked to see sample content. 6. On a selected workout: `get_workout_details`, then `get_workout_playlist` if they need segments or media. 7. Call `get_embed_code_samples` only if they asked for copy-paste code after the plan. 8. Reply in 3 short bullets: recommended path, what to copy next, one follow-up ask. ## Output - Integration path name (embed / stream / custom player) - One primary snippet with placeholders only - Suggested next tool or question ## Guardrails - Never invent or print live API keys. Use `YOUR_ORG_ID` and `YOUR_API_KEY`. - No pricing, checkout, or plan catalogs. - Prefer guest-safe tools when the user is not connected. - See [references/integration-paths.md](references/integration-paths.md) and [references/guardrails.md](references/guardrails.md).
Referenced files: 6
Package details
Publisher declarations from the archived package. These are separate from our research and the live service's terms.
- Package author
- Hyperhuman
Package observed Oct 2, 2026.
Technical details
- First seen
- Sep 30, 2026 · 22:02 UTC
- Last seen
- Oct 2, 2026 · 00:00 UTC
- Collection status
- Collected
plugin_asdk_app_69a6a3c8b4708191bdb5d0d8141c2176
Download plugin data (JSON)