agentleFS
rrudolrrudol/frisco/.agents/skills/meal-planner/SKILL.md

meal-planner

Skill8 starsChanged 8 months agoMITView on GitHub

Plan meals, search Frisco for matching products, build a cost-effective cart with nutrition tracking, and iterate with the user to optimize.

8 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

12 rules. 0 are common, 9 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 7 edited versions we read keep these.

  • 9

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If binary missing: make build.

    line 13
  2. Very specific:

    Always use --category-id to narrow results.

    line 28
  3. Distinctive:

    Use subagents in parallel to search multiple product groups at once (e.g., meat, dairy, vegetables, pantry).

    line 44
  4. Distinctive:

    When a product is unavailable, search for alternatives and propose a substitution — update the meal plan accordingly.

    line 46
  5. Very specific:

    Wait for explicit confirmation: before running without --dry-run.

    line 53
  6. Distinctive:

    Add a macro summary table to each day in the plan.

    line 65
  7. Distinctive:

    Add a weekly average at the bottom.

    line 66
  8. Distinctive:

    After the cart is filled, review it with the user:

    line 81
  9. Distinctive:

    After every change, show the updated cart total.

    line 105
  10. Distinctive:

    Substitutions: update recipe text when swapping products.

    line 117
  11. Distinctive:

    After the user places an order, review what was actually ordered:

    line 125
  12. Distinctive:

    Use this to verify the order matches the plan and to inform future planning sessions.

    line 138

“Common” is measured: kept by at least 6 in 10 of the 7 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Frisco Meal Planner

Help the user go from meal ideas or dietary guidelines to a filled Frisco.pl cart with full nutritional breakdown.

Prerequisites

  1. ./bin/frisco session verify — if it fails, guide the user through session login or session from-curl.
  2. If binary missing: make build.
  3. Gather context from the user:
    • Dietary guidelines, restrictions, or just meal ideas.
    • How many people and which days (e.g., "weekend for 2").
    • Budget expectations (if any).
    • What they already have at home (oils, spices, staples) — ask early to avoid buying things they don't need.

Phase 1: Meal Plan

  • Build a plan matching the user's input (number of days, meals per day, portions).
  • Save to a markdown file (not in the repo — ask user where they want it, e.g., Obsidian vault).
  • The plan is a living document — update it whenever the cart changes.

Phase 2: Product Search

Always use --category-id to narrow results. Without it, searches return irrelevant matches. Full category tree is in categories.md at the project root.

# Best-match pick (scores by name match + price/kg + pack size):
./bin/frisco products pick --search "PHRASE" --category-id CATID --top 3

# Raw search when pick isn't enough:
./bin/frisco products search --search "PHRASE" --category-id CATID --format json 2>/dev/null | \
  jq '[.products[] | select(.product.isAvailable == true) |
    {id: .productId, name: .product.name.pl, price: .product.price.price,
     grammage: .product.grammage, unit: .product.unitOfMeasure}] | .[0:5]'

# Add single product by search:
./bin/frisco cart add --search "PHRASE" --category-id CATID --quantity N

Use subagents in parallel to search multiple product groups at once (e.g., meat, dairy, vegetables, pantry). Each subagent uses the Go CLI + jq.

When a product is unavailable, search for alternatives and propose a substitution — update the meal plan accordingly.

Phase 3: Cart Assembly

  1. Collect all product IDs and quantities into a JSON file.
  2. Dry-run first: ./bin/frisco cart add-batch --file FILE --dry-run
  3. Show the user a full summary table with names, prices, and quantities.
  4. Wait for explicit confirmation before running without --dry-run.

Phase 4: Nutrition / Macros

The user cares about nutritional values. For every meal plan, provide per-meal and per-day macro breakdowns (kcal, protein, fat, carbs).

Data sources (in priority order):

  1. Frisco API — ./bin/frisco products nutrition --product-id PID --format json returns per-100g values for many products. Use subagents to batch-query nutrition for all key ingredients.
  2. Standard nutritional databases — for products where the API returns no data, use well-known reference values. Mark these clearly as "estimated" in the plan.

How to present:

  • Add a macro summary table to each day in the plan.
  • Add a weekly average at the bottom.
  • Note which values come from the API vs estimates.
  • Flag any days that seem too high or too low for the user's goals.
# Query nutrition for a product:
./bin/frisco products nutrition --product-id PID --format json 2>/dev/null | \
  jq '{kcal: (.nutrition.nutriments[] | select(.name.pl | test("kcal")) | .in100Gramms),
       protein: (.nutrition.nutriments[] | select(.name.pl | test("Białko")) | .in100Gramms),
       fat: (.nutrition.nutriments[] | select(.name.pl | test("^Tłuszcz")) | .in100Gramms),
       carbs: (.nutrition.nutriments[] | select(.name.pl | test("^Węglowodany")) | .in100Gramms)}'

Phase 5: Cost Optimization

After the cart is filled, review it with the user:

# Most expensive items:
./bin/frisco cart show --sort-by total --top 20

# Or via JSON for detailed analysis:
./bin/frisco cart show --format json 2>/dev/null | \
  jq '[.products[] | select(.quantity > 0) |
    {name: (.product.name.pl // .product.name.en), price: .product.price.price,
     qty: .quantity, total: (.product.price.price * .quantity), grammage: .product.grammage}] |
  sort_by(-.total) | .[0:20]'

Propose savings by working through these strategies:

  1. Things the user already has — pantry staples accumulate. Ask.
  2. Frozen vs fresh — seasonal pricing matters. Frozen fruits/vegetables can be much cheaper.
  3. Waste reduction — if a package is much larger than needed, either spread it across multiple days in the plan or find a smaller alternative. Fewer product types with full utilization beats variety that rots.
  4. Duplicates — parallel searches can pick different products for the same ingredient.
  5. Quantity right-sizing — don't order 2 packs when 1 covers the week.
  6. Cheaper alternatives — different brands, non-BIO, different cuts, store brand.
  7. Better bought elsewhere — some niche items are overpriced on Frisco vs pharmacies or specialty stores. Flag these.

After every change, show the updated cart total.

./bin/frisco cart remove --product-id PID
./bin/frisco cart add --product-id PID --quantity N
./bin/frisco cart remove-batch --product-ids PID1,PID2,PID3

Phase 6: Plan Reconciliation

The meal plan must always reflect what's actually in the cart:

  • Substitutions: update recipe text when swapping products.
  • Removed items: rework affected meals to use remaining ingredients.
  • Portion changes: recalculate when some days serve more people.
  • Ingredient budget: for items with large packages, add a table showing which days use what amount, so nothing goes to waste.
  • Macros: recalculate nutrition whenever ingredients change.

Phase 7: Order Review

After the user places an order, review what was actually ordered:

# List recent orders:
./bin/frisco account orders list

# See products in a specific order:
./bin/frisco account orders products --order-id ORDER_ID --sort-by total

# Full order details:
./bin/frisco account orders get --order-id ORDER_ID

Use this to verify the order matches the plan and to inform future planning sessions.

Principles

  • Use the Go CLI — products pick, products search, cart add --search, cart show --sort-by. No external scripts.
  • Confirm before cart changes — always show what will happen and wait for OK.
  • Minimize waste — adapt the plan to use what's bought in full.
  • Respect dietary restrictions — never substitute excluded ingredients, even if cheaper.
  • Track nutrition — the user wants to know what they're eating. Include macros in every plan.
  • Iterate — present, discuss, adjust. Don't aim for perfection in one pass.
  • Show the running total — the user should always know where they stand.
  • Keep personal files out of the repo — plans, shopping lists, cart JSONs go where the user wants them (e.g., Obsidian vault), not in the git repo.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.