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 skeletonarmytech/marrow-mcp --skill marrow-healthgit clone --depth 1 https://github.com/skeletonarmytech/marrow-mcpWrote 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/skeletonarmytech/marrow-mcp/marrow-health)<a href="https://agentmods.dev/skills/skeletonarmytech/marrow-mcp/marrow-health"><img src="https://agentmods.dev/badge/skills/skeletonarmytech/marrow-mcp/marrow-health/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/skeletonarmytech/marrow-mcp/marrow-health"><img src="https://agentmods.dev/badge/skills/skeletonarmytech/marrow-mcp/marrow-health.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 | $0.00077 | $0.00592 |
| Opus 5.5 | $0.00031 | $0.00237 |
| Sonnet 5.5 | $0.00015 | $0.00118 |
| Haiku 4.5 | $0.00008 | $0.00059 |
Grade A, and why
marrow-health 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 — 49 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Marrow Health
The marrow MCP server is the user's own Marrow mirror, synced from their
iPhone. Everything it returns is their data, reconciled the way the app
shows it. Answer from it, never from assumptions.
Which tool answers what
- "How am I doing" or anything about today:
daily_brief. Each key metric comes with a phrase against the user's own trailing normal (normal for you, above your normal, well below your normal), food and water against targets, last night's sleep, training done, mood, symptoms and flags. Start here before giving any advice. - Is a number high or low for this person:
baselines(mean, spread, z, phrase, 7-day mean, 90-day trend). Use its phrases, not population ranges. - What is different lately, weekly reviews:
what_changed(recent days against the period before, notable items first, food, training, sleep, bedtime shift, streams that started or stopped). - Before concluding something did not happen:
data_quality. A stale or missing stream is not a zero. - Raw detail:
health_summary(a plain day table),metric_daily,metric_samples(page withbefore),workouts,nutrition_log,strength_log,sleep_log(the only correct source for nights),symptoms_and_mood,targets. - Every read takes
daysback orfromandtoon the phone's calendar (YYYY-MM-DD).
Logging
Only when the user asks. log_meal, log_water, log_weight,
log_workout, log_strength, log_sleep, log_symptom, log_mood,
edit_entry, delete_entry, set_targets, undo_entry. The phone
applies each change and writes it to Apple Health. If a write answers that
write-back is turned off, tell the user where the switch is (Marrow,
Export, Connect your AI, Allow AI write). Never log, change or remove
anything the user did not ask for. Confirm what was logged in one line.
Wording
Neutral wellness wording. Describe, compare with the user's own normal, suggest. Do not diagnose or alarm. Quote units. Say plainly when data is missing or stale.
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 First seen · 49 lines · 77 tokens per session scan A 719f3b74f00c
marrow-health is a skill published in the GitHub repository skeletonarmytech/marrow-mcp (3 stars, last pushed yesterday), licensed MIT. It adds 77 tokens to every session and 592 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-10-22.
Other skills, from other repositories
mirobody
Self-host Mirobody, the open-source health data engine, and use it from an agent. Use when the user wants to run a personal or family health record on their own machine, bring lab reports and wearables (Apple Health, Garmin, Oura, WHOOP) into one place an AI can read, expose their health data to Claude Code, Codex…
mitiq-error-mitigation
A local experiment runner for testing four ways to reduce errors in sampled quantum circuits: zero-noise extrapolation, readout correction, probabilistic error cancellation, and Clifford data regression. It compares noisy and mitigated results under the same sampling budget.
quantum-ground-state
A narrowly defined quantum solver for the ground-state energy of a two-quantum-bit Pauli Hamiltonian in a specified one-excitation sector. It uses a noiseless statevector VQE and checks the result against an exact calculation; VQE is a method that searches for a low-energy quantum state.
qmclaw-workbench
A local simulator for superconducting quantum-bit experiments, including measurements used to study and tune a qubit. It produces bounded synthetic results without connecting to laboratory hardware or quantum cloud services.
quantum-advection
A local workflow for simulating one-dimensional advection, the movement of a quantity through space, with a quantum circuit. It supports constant speed, periodic boundaries, and a finite-difference approximation.
quantum-amplitude-amplification
A local Qiskit workflow for amplitude amplification, a quantum method that increases the chance of measuring selected answers. It can explain the method, build a circuit, or run calculations on a simulator.