Agent Chef

Propose this week's candidates

propose_recipes

Destructive

Store candidate dinner recipes for the week and generate numbered options, a ballot with a vote link placeholder, and each member's voting link. Re-calling replaces the week's candidates.

Instructions

Start a weekly cycle: store the candidate recipes (normally 10) and return numbered candidates, a ballot_text (with a {vote_link} placeholder for message_group), and each member's vote link. Re-calling for the same week_start replaces the candidates.

Input Schema

TableJSON Schema

Name	Required	Description	Default
`recipes`	Yes
`week_start`	No	YYYY-MM-DD, defaults to today

Output Schema

TableJSON Schema

Name	Required	Description
`week`	Yes
`similar`	Yes	New recipes that closely resemble one the household already has; consider using theirs
`replaced`	Yes	New recipes swapped for the household's own version of the same dish; the existing recipe is on the ballot instead
`candidates`	Yes
`image_urls`	Yes
`vote_links`	Yes
`ballot_text`	Yes	Message text with a {vote_link} placeholder

Schema Changelog

Changes observed during successful MCP inspections.

Changed10 schema fields changedv0.1.1Oct 3, 2026

changedInput schema / properties / recipes / items / properties / image_url / description

Previous value: -"A real, publicly reachable photo of the dish: the og:image of the recipe page you used, or one found via web/image search. The server verifies the URL is an image and drops it otherwise. Never invent URLs; prefer import_recipe, which captures the source photo."New value: +"A real, publicly reachable photo of the dish: the og:image of the recipe page you used, or one found via web/image search. The server verifies the URL is an image and drops it otherwise. Never invent URLs; prefer import_recipe, which captures the source photo. For generated images use set_recipe_image afterwards."

addedOutput schema / properties / candidates / items / properties / admin_pick

Added value: +{
+  "description": "Selected even though the vote alone would not have picked it: chosen by an admin, or added off the ballot",
+  "type": [
+    "boolean",
+    "null"
+  ]
+}

addedOutput schema / properties / candidates / items / properties / image_generated

Added value: +{
+  "description": "Photo was generated by Agent Chef, so it is safe to show publicly",
+  "type": [
+    "boolean",
+    "null"
+  ]
+}

addedOutput schema / properties / candidates / items / properties / is_public
```
Added value: +{
+  "type": [
+    "boolean",
+    "null"
+  ]
+}
```

addedOutput schema / properties / candidates / items / properties / public_url

Added value: +{
+  "description": "Indexable public page, when published",
+  "type": [
+    "string",
+    "null"
+  ]
+}

addedOutput schema / properties / candidates / items / properties / source_recipe_id

Added value: +{
+  "description": "Set when this is the household's copy of a Community recipe",
+  "type": [
+    "string",
+    "null"
+  ]
+}

addedOutput schema / properties / replaced

Added value: +{
+  "description": "New recipes swapped for the household's own version of the same dish; the existing recipe is on the ballot instead",
+  "items": {
+    "additionalProperties": true,
+    "properties": {
+      "existing_title": {
+        "$ref": "#/properties/week/properties/id"
+      },
+      "recipe_id": {
+        "$ref": "#/properties/week/properties/id"
+      },
+      "title": {
+        "$ref": "#/properties/week/properties/id"
+      }
+    },
+    "required": [
+      "title",
+      "recipe_id",
+      "existing_title"
+    ],
+    "type": "object"
+  },
+  "type": "array"
+}

addedOutput schema / properties / similar

Added value: +{
+  "description": "New recipes that closely resemble one the household already has; consider using theirs",
+  "items": {
+    "additionalProperties": true,
+    "properties": {
+      "recipe_id": {
+        "$ref": "#/properties/week/properties/id"
+      },
+      "similar_to": {
+        "items": {
+          "additionalProperties": true,
+          "properties": {
+            "recipe_id": {
+              "$ref": "#/properties/week/properties/id"
+            },
+            "title": {
+              "$ref": "#/properties/week/properties/id"
+            }
+          },
+          "required": [
+            "recipe_id",
+            "title"
+          ],
+          "type": "object"
+        },
+        "type": "array"
+      },
+      "title": {
+        "$ref": "#/properties/week/properties/id"
+      }
+    },
+    "required": [
+      "recipe_id",
+      "title",
+      "similar_to"
+    ],
+    "type": "object"
+  },
+  "type": "array"
+}

addedOutput schema / properties / week / properties / share_url

Added value: +{
+  "description": "Public page showing this week's winners",
+  "type": [
+    "string",
+    "null"
+  ]
+}

changedOutput schema / required

Previous value: -[
-  "week",
-  "candidates",
-  "ballot_text",
-  "image_urls",
-  "vote_links"
-]New value: +[
+  "week",
+  "candidates",
+  "replaced",
+  "similar",
+  "ballot_text",
+  "image_urls",
+  "vote_links"
+]

First observedv0.1.0Sep 7, 2026

TDQS

A3.8/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare destructiveHint=true and openWorldHint=true. The description adds valuable context beyond that: it specifies the destructive behavior ('Re-calling for the same week_start replaces the candidates'), the expected volume ('normally 10'), and the return structure. This goes beyond the annotations, meriting a 4 rather than the baseline 3.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Three sentences with no waste, front-loading the core action and immediately following with return values and the replacement behavior. Every sentence earns its place.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the complexity, schema richness, and presence of an output schema, the description provides what's needed: the action, return format, and destructive replacement behavior. It's complete enough for an agent to call correctly, though it could clarify prerequisites or alternatives.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 50%, with the recipes array and its nested properties well-documented. The description adds context about week_start defaulting to today and the typical number of recipes, but doesn't fully compensate for the undocumented half of the parameter details.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb+resource: store candidate recipes and return numbered candidates, ballot_text, and vote links. It's clear what the tool does, though it doesn't explicitly distinguish itself from sibling add_candidates or set_this_weeks_recipes, so it falls short of a 5.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description implies usage as the start of a weekly cycle, giving context, but it doesn't explicitly say when to use this versus add_candidates or set_this_weeks_recipes. No exclusions or prerequisites are stated.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.