Review an Analyze agent prompt

POST

/api/v3/agents/analyze/prompt-review

curl --request POST \
  --url https://api.tryletterhead.com/api/v3/agents/analyze/prompt-review \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{ "api": true, "prompt": "Tell me which newsletter made the most revenue last month.", "scope": [ "the-daily" ] }'

• Production

Review an Analyze agent’s prompt before activation and return a structured assessment — whether it’s fully in scope, specific issues to fix, any out-of-scope asks, and a suggested rewrite.

Authorizations

• bearerAuth

Request Body ^required

Media type application/json

object

api

required

Required. Set to true to authenticate with an API key.

boolean

prompt

required

The prompt to review, up to 1500 characters.

string

scope

Optional list of newsletters the agent is scoped to, for context.

Array<string>

Example

{
  "api": true,
  "prompt": "Tell me which newsletter made the most revenue last month.",
  "scope": [
    "the-daily"
  ]
}

Responses

200

200 OK

Media type application/json

object

items

object

compatible

boolean

issues

Array<object>

object

concern

string

fix

string

outOfScopeFeatures

Array<string>

suggestedPrompt

string

suggestedRewrite

string

nullable

message

string

total

integer

Example

{
  "items": {
    "compatible": false,
    "issues": [
      {
        "concern": "The prompt asks about revenue, which the agent can't see.",
        "fix": "Ask about engagement metrics instead, like open or click rate."
      }
    ],
    "outOfScopeFeatures": [
      "revenue analysis"
    ],
    "suggestedPrompt": "Tell me which newsletter had the highest open rate last month.",
    "suggestedRewrite": "Tell me which newsletter had the highest open rate last month."
  },
  "message": "",
  "total": 1
}