meal-planning
Skill2 starsChanged 5 months agoMITView on GitHub
Plan meals collaboratively in chat, then publish the finalized plan to mealforge via its MCP tools. Use when the user wants to plan meals (a week of dinners, a few breakfasts, a single meal), revise a pushed plan, switch or complete the active plan, or push a plan to mealforge.
3 other projects use a version of this.
Adopted by
Most recently taken up by
- matthull/my-skillstheir own version, first seen Oct 20, 2026
- cooklang/cooklang-skillstheir own version, first seen Oct 13, 2026
- cook-md/cook-mcptheir own version, first seen Oct 8, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- cooklang/cooklang-skills★ 15same name
- cook-md/cook-mcp★ 1same name
- matthull/my-skills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository5 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- feature-developmentSkill
- tanstack-router-patternsSkill
- trpc-drizzle-patternsSkill
The file
We split this skill into the rules it gives an agent
2 rules. 0 are common, 1 is distinctive, and 1 is very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
1
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 32
If the user wants something they've had before ("one of our favorites", "that carnitas"), find it with list_favorites or search_recipes, …
- Distinctive: line 41
Wait for an explicit instruction ("push it", "send it over", "that's the plan").
“Common” is measured: kept by at least 6 in 10 of the 3 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.
Files
Meal planning with mealforge
Plan meals collaboratively in chat, then publish the finalized plan to the mealforge app. A plan is any set of 1+ meals — a full week of dinners, a weekend of brunches, breakfast-lunch-dinner-snacks, or one special meal. Each meal is typed breakfast, lunch, dinner, or snack. Planning is conversation — as much back-and-forth as it takes. Publishing is a command — it happens exactly once, only when the user says so, and always via the two-step flow: save each recipe with create_recipe, then assemble the plan with push_meal_plan using recipeIds only.
The plan lifecycle
Plans move through three states, and the household cooks from exactly one active plan at a time:
- upcoming — pushed but not started; plans queue here.
- active — the plan being cooked from right now. Its grocery list is the one in the app. A brand-new push becomes active automatically only when nothing else is active.
- completed — cooked through; browsable history.
Lifecycle moves are user commands, not your judgment calls: complete_meal_plan when they say they're done ("we finished the week", "wrap up this plan"), activate_meal_plan when they pick what's next. Activation fails while another plan is active — complete that one first (confirm with the user).
When to use
- The user wants to plan meals ("let's plan next week", "plan me three breakfasts", "meal plan time")
- The user revises a pushed plan ("swap the salmon", "add a couple of snacks to the plan")
- The user tells you to push/send the plan to mealforge
- The user wants to switch plans, finish a plan, or bring back a favorite plan
How to apply
1. Gather context before proposing anything
get_active_meal_planandlist_meal_plans(limit 6) — see what's cooking now and what was cooked recently; don't repeat recent meals unless asked. Each meal'scookedAt(null = not cooked yet) is the household's check-off, so you can answer "what's left to cook this week?".- If the user wants something they've had before ("one of our favorites", "that carnitas"), find it with
list_favoritesorsearch_recipes, confirm withget_recipe, and remember its recipeId — reuse the id at push time instead of recreating the recipe. - For "one of our favorite plans", use
list_meal_planswithfavoritesOnly: trueand rebuild from its recipeIds.
2. Draft the whole plan at once
Ask (or infer) the plan's shape first: how many meals, which meal types. From whatever cues the user gives ("lots of crock pot, salmon once, use the pork shoulder"), propose the complete plan in one reply — meal type, meal name, one line each. No full recipes in chat unless asked. Then iterate meal-by-meal as the user reacts. Stay in conversation as long as it takes; never call any write tool during drafting.
3. Publish — only on command, always in two steps
Wait for an explicit instruction ("push it", "send it over", "that's the plan"). Never push drafts, never push "to save progress". Then:
Step A — save each NEW recipe with create_recipe, one call per recipe. Small flat payloads are reliable; one giant nested payload is not. Each call returns a recipeId — collect them. Recipes the user is repeating from history already have ids; don't recreate those.
Step B — one push_meal_plan call with ids only:
{"name":"Slow Cooker Week","meals":[
{"mealType":"dinner","recipeId":12},
{"mealType":"dinner","recipeId":13},
{"mealType":"breakfast","recipeId":9}
]}name is optional but nice — unnamed plans show as "Meal Plan {id}", and a plan must be named before the household can favorite it in the app. The result tells you the planId and whether the plan landed active or upcoming — relay that ("it's live now" vs "it's queued behind the current plan") and share the app URL from the tool result.
4. Recipe data quality
The grocery list is derived automatically from ingredients, so precision here is the whole ballgame:
title,stepsMarkdown(numbered list:"1. …\n2. …", real cookable steps), andingredientsare required. Includedescription,servings,prepMinutes,cookMinutes,tags, andmealTypes(which slots the recipe suits, e.g.["breakfast","lunch"]— planning it as a meal also tags it automatically).quantity: a number (1.5, or "1/2" is accepted). Omit or null only for to-taste items (salt, pepper).unit: "lb", "oz", "g", "kg", "cup", "tbsp", "tsp", "clove", "can", "bunch" — or null for countable items ("3 zucchini").section: one ofproduce,meat-seafood,dairy-eggs,bakery,pantry,frozen,spices,beverages,other.- Name ingredients identically across recipes — "yellow onion" everywhere, not "onion" in one recipe and "yellow onion" in another — so the grocery list merges them into one line.
5. If a tool call fails
The error names the exact fields and shows a valid example. Fix only those fields and retry once. Don't restructure the whole payload, don't switch tools, don't retry more than twice — if it still fails, tell the user what the error said.
6. Revising a pushed plan
Adjust in chat, then — again only when told — re-push with the same planId: create_recipe for any newly added meals, then push_meal_plan with planId and the full plan's ids (unchanged meals keep their existing recipeIds). The app replaces the meals and regenerates the grocery list; checked-off grocery items that didn't change stay checked, and meals that survive the revision (same recipeId and mealType) keep their cooked check-offs, so mid-plan re-pushes are safe. Never omit planId when revising — that would create a duplicate plan.
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
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.

