CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

rampwise-report

Daily, weekly and monthly training reports from the user's own Garmin data, as Markdown with figures and as an interactive single-file HTML page (works offline, light and dark mode, period buttons, hover tooltips). Use for "give me a report", "daily/weekly/monthly summary", "show my progress", "dashboard", "how did my week go", "how am I doing against my goal", and for sharing or archiving a training review. Includes recovery, week-over-week comparison, injury-risk watch, plan against what was done, pace/speed/efficiency trends, strength logs, goal meters and tomorrow's adapted session. Goals and units follow the user's own settings; nothing is compared with other people. Uses rampwise-analyze, rampwise-adapt and rampwise-recovery for the numbers.

Quality

97%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Paths. <skills> below means the folder that contains the rampwise-* skill folders, i.e. the parent of this skill's own folder. Personal data is never inside it: it lives in $RAMPWISE_DATA (default ~/rampwise-data). Reports are written to the output folder, never into the plugin folder.

Garmin Report (daily, weekly, monthly)

Question you answer: what actually happened, and is it moving toward the user's goals? One command builds the report; you then read it with the user and explain what stands out.

Run

python -X utf8 "<skills>/rampwise-report/scripts/report.py" daily
python -X utf8 "<skills>/rampwise-report/scripts/report.py" weekly
python -X utf8 "<skills>/rampwise-report/scripts/report.py" monthly --format html

Options: --date YYYY-MM-DD (report as of that day), --format md|html|both (default both), --out DIR, --no-deep (skip the FIT-file analysis of the latest activity), --no-adapt (skip tomorrow's adaptation). A weekly report takes about half a minute because it reads Garmin live.

Output: ~/rampwise-coach/reports/<date>-<mode>/ (the output folder; --here uses ./rampwise-coach, preferences.output_dir or RAMPWISE_OUT change it) with

  • report.md plus figures/*.svg (Markdown with real figures; the SVGs carry their own colours and dark-mode rules),
  • report.html (single file, no network, no scripts: period buttons are pure CSS, charts show values on hover, collapsible sections, prints cleanly),
  • reports/latest-<mode>.html, always the newest one (bookmark it).

Tell the user where the files are; if they want to see the HTML in the chat, publish it as an Artifact or open it in a browser. Do not paste raw health data anywhere outside their machine without asking.

What is in each report

SectionDailyWeeklyMonthly
What stands out (readiness, watch level, biggest changes)yesyesyes
Recovery: HRV against the normal band, resting HR, sleep14 or 28 days14 or 28 days14 or 28 days
Injury-risk watch: flags with evidence labels, questionsyesyesyes
This week against last week (rolling 7 days; calendar week from Wednesday)yesyesyes
Minutes per day by sport14 or 28 days14 or 28 days14 or 28 days
Latest activity against the last similar one, aerobic readingyes
Recent activities table (14 days), each against the last similar oneyesyes
Tomorrow, adapted (rampwise-adapt)yes
Minutes per week by sport; consistency heatmap4/8/12 weeks12 weeks, heatmap
Running pace, run efficiency, longest run; riding speed, longest ride4/8/12 weeks4/8/12 weeks
Strength sessions per week, top set per exercise (first against latest)yesyes
Weekly recovery means (HRV, resting HR)yesyes
Plan against done, next 7 days of the planyesyes
Goal metersyesyesyes

Sections appear only when there is data for them (no cycling, no cycling charts). Running is shown as pace (min/km) and cycling as km/h unless preferences.units in athlete.json says otherwise.

Goals (meters)

A meter shows the start value, the latest reading and the target. The start is the goal's recorded baseline if there is one, else the first reading in the last 12 weeks; "now" is the best reading of the last 28 days. Supported automatically: a ride-speed goal (target_speed_kmh, flat rides of at least half the target distance), a running-pace goal (target_pace_s_per_km) and tracked goals in athlete.json (goals[].track = {metric, target, unit} with metric weekly_minutes, long_run_km, long_ride_km, training_days_per_week or strength_sessions_per_week). No goal set means no meters: say so and offer to add one with athlete.py goal-add. A meter is a distance to the goal, never a forecast. No paper read gives a rate of improvement for one person; never write "you will reach it by ...".

How to present it

  1. Read the "What stands out" list and the injury-risk watch first. If the watch is AMBER or RED or readiness is RED, say so before any good news.
  2. Pick 2 or 3 findings that the user can act on (a drop or a jump, a trend, a flag). Quote the figure, not a verdict.
  3. Efficiency, pace and speed move with heat, hills, wind, illness and the easy/hard mix: say "compare similar sessions" instead of "you got fitter".
  4. Ask the questions the report lists (pain, how they feel). Record answers (athlete.py check-in).
  5. Offer the next step: rampwise-adapt for tomorrow, rampwise-planner for the week. Keep the chat answer short; the report holds the detail.

Design notes (references/design.md)

Fixed colours per sport (run blue, ride orange, strength green, neutral grey), thin marks, a legend whenever there are two or more series, values on hover (HTML) and comparison tables (both formats), light and dark palettes, and no JavaScript.

Limits

  • Charts are as good as the data: missing nights leave gaps, a watch without heart-rate data has no efficiency chart.
  • The report states "not medical advice"; flags come from data and the user's own reports.
  • Strength top sets need sets saved on the watch (strength_history.py reads them).

Papers

The studies behind this skill are in references/papers/ (INDEX.md lists the read depth). Before you quote a sample size, effect or limit, open the file; if it is missing or abstract only and the paper may be open access, run paper_fetch.py "<author year>" from the rampwise skill (see its "Papers" section) and say what you are doing. Paywalled papers stay abstract only: say so and claim no more.

Repository

Azahir21/rampwise

Path

skills/rampwise-report/SKILL.md

Last updated
First committed

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.