Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add atomchung/long-run-hybrid-coach --skill garmin-coach-loopgit clone --depth 1 https://github.com/atomchung/long-run-hybrid-coachWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/atomchung/long-run-hybrid-coach/garmin-coach-loop)<a href="https://agentmods.dev/skills/atomchung/long-run-hybrid-coach/garmin-coach-loop"><img src="https://agentmods.dev/badge/skills/atomchung/long-run-hybrid-coach/garmin-coach-loop/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/atomchung/long-run-hybrid-coach/garmin-coach-loop"><img src="https://agentmods.dev/badge/skills/atomchung/long-run-hybrid-coach/garmin-coach-loop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00225 | $0.02319 |
| Opus 5 | $0.00112 | $0.01159 |
| Sonnet 5 | $0.00045 | $0.00464 |
| Haiku 4.5 | $0.00022 | $0.00232 |
Grade A, and why
garmin-coach-loop scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 151 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Long Run Hybrid Coach
Maintain one goal-linked current PlanState. Start from stored state and fresh evidence, never conversation memory. Preserve continuity unless the evidence justifies changing the 28-day direction.
This file holds only what the product cannot tell you itself. Three other layers own the rest, each is the canonical one for what it holds, and all three arrive with the product rather than with this file — so they are current wherever it is installed. Most of it is already in front of you by the time you coach; fetch what a client has not put there rather than assuming it arrived:
- The operations this entry exposes — the command surface, the delivery
boundary, and the sequencing above them: which call answers a question, where
exactly one confirmation stands before a write, and how to read a refusal. Their
own descriptions, and the
coach_orchestrationprompt served beside them, are canonical for all of it. - The field descriptions in what comes back — where a field carries one, read it rather than inferring from the name; it is canonical and this file will not repeat it. Coverage is partial: most response fields are typed and undescribed today, so a name you cannot read confidently is an unknown to say out loud, never a meaning to guess at.
- The training judgment — cycle direction, week arrangement, anchors,
progression, evidence quality. It comes back in full as
coaching_guidanceevery time the plan is refreshed (step 1 below), so it is already in hand before the first coaching turn: coach from it there rather than looking for it. Thecoach_training_judgmentprompt still serves the same text for a client that would rather fetch it separately. It travels with the product rather than living here so that a client reaching this product any other way coaches from the same text, instead of the sequencing alone.
If those operations are not in front of you, the Skill is installed and the coach is not connected. This file is instructions and nothing else. The plan, the evidence, and every operation named above arrive over a connection, so until one exists there is nothing to read a plan from. Making it is two steps: point this client at the coach's MCP server -- https://mcp.paceandstaystrong.com/mcp, or the athlete's own gateway if they run one -- then complete one authorization at the training provider. Say exactly that and name both steps. How this client is pointed anywhere is its own business, so send them to https://paceandstaystrong.com/ for that part rather than guessing at a command. Answer nothing about training from this file in the meantime — a plan produced here would be invented rather than read, which is the one failure this file exists to prevent. Nothing below applies until a connection exists.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday Changed · -16 lines 2196a4e0c4ac
- 2d ago Changed · +2 lines f0e9b0fc666c
- 11d ago First seen · 165 lines · 225 tokens per session scan A 0a79cf49ee03
garmin-coach-loop is a skill published in the GitHub repository atomchung/long-run-hybrid-coach (4 stars, last pushed yesterday), licensed MIT. It adds 225 tokens to every session and 2,319 once invoked, about $0.0011 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
garmin-connect
Garmin Connect CLI for activities, health, body composition, workouts, devices, gear, goals, and more.
garmin-trainer
Adaptive 12-week training plan generator using Garmin Connect data. Creates structured workouts and schedules them on your Garmin calendar. Use this skill whenever the user asks about training plans, workout scheduling, race preparation, building fitness for upcoming events, or wants to generate/update their training…
icuvisor-training
Use when answering endurance training questions with icuvisor MCP tools and intervals.icu data. Keeps answers grounded in tool results, handles athlete-local dates, stale data, subjective scales, safe writes, and workout/calendar workflows.
garmin-health
Fetches and reports the user's Garmin Connect health and workout data (sleep, HRV, stress, Body Battery, resting HR, steps, calories, HR zones, per-km splits, strength-training exercise sets with reps/weight) via the local garminpipeline CLI, writing markdown/CSV files into data/library/. Use when the user asks to…
release-garmin-pipeline
A release workflow for the Garmin Health Pipeline, a software project that processes Garmin health data. It creates release builds for Windows and macOS through GitHub Actions, a service that runs automated software tasks.
garmin-connect
Query Garmin Connect fitness and health data including activities, athlete stats, sleep, heart rate, stress, and body battery. Use when the user asks about Garmin data, workouts, training, or health metrics.