What endpoints does the Rypl API have?
Updated
The Rypl API has endpoints for your channels, comments, direct messages, posts and ads, your agent's replies, counts for a period, the Needs attention list, your agents and your Global rules, plus actions to hide, unhide and mark a comment as reviewed and to change your Global moderation rules. It reads with GET, changes with POST and PATCH, and answers in JSON.
Basics
| Address | https://app.rypl.is/api/v1 |
| Key | Authorization: Bearer rypl_… with every request |
| Limits | 120 requests a minute per connection, 30 changes a minute. Over that: 429 with Retry-After |
| Times | ISO 8601: 2026-11-01 or 2026-11-01T09:30:00Z. since and until are inclusive |
| Lists | Newest first, as { "data": [...], "next_cursor": "…" }. Send cursor=<next_cursor> for the next page; null means the end |
| Ids | Channels, authors and posts keep their Facebook, Instagram or TikTok ids. A row's id is Rypl's own, and is the one actions take |
The data can be a few seconds behind your Inbox. Errors are listed in Why can't my app connect to Rypl?
Endpoints
| Endpoint | What it returns | Needs |
|---|---|---|
GET /me | Whose key this is, what it may read, its channels and when it ends | Any key |
GET /channels | Your connected channels, with their settings if allowed | Any key |
GET /comments | Comments on your posts and ads | Comments |
GET /messages | Direct messages on Facebook Messenger and Instagram | Direct messages |
GET /posts | Posts and ads people commented on, with counts | Comments |
GET /replies | Your agent's replies, each with what it answered | Comments or Direct messages |
GET /summary | Counts for a period, optionally grouped | Comments or Direct messages |
GET /attention | The Needs attention list | Comments or Direct messages |
GET /agents | Your agents: instructions, knowledge, reply rules | Settings |
GET /rules | Your Global rules: rule cards and banned words | Settings |
POST /comments/{id}/hide | Hides a comment on its platform | Moderate comments |
POST /comments/{id}/unhide | Shows a hidden comment again | Moderate comments |
POST /comments/{id}/review | Marks a comment as reviewed; {"undo": true} takes it back | Moderate comments |
POST /rules/cards | Adds a rule to your Global rules | Change moderation rules |
PATCH /rules/cards/{id} | Changes a rule's text or action, or switches it off or on | Change moderation rules |
PATCH /rules | Changes the mode, the default action or the banned words | Change moderation rules |
Filters for /comments and /messages
| Parameter | Values |
|---|---|
period, timezone | today, yesterday, last_7_days, last_30_days; timezone is a name such as Europe/Berlin (UTC without it). Not together with since and until |
since, until | When the row was written on the platform |
updated_since | Rows that changed in Rypl since then: analysed, replied to, hidden or deleted |
channel, platform | A channel id from /channels; facebook, instagram, tiktok |
view | An Inbox tab: needs_review, reply_needed, negative, ai_replies, hidden_or_deleted, unanswered_questions |
sentiment, urgency, intent | For example negative, high, complaint |
language | A language code, for example en |
is_question, replied, hidden, deleted, from_page | true or false. from_page=false keeps only customers' rows |
post, parent, author | A post id; a comment's id, for its replies; an author id |
author_name | Part of a name or username, 2 to 60 characters |
search, search_any | Text or a name, up to 100 characters; or up to 10 words or phrases, separated by commas, matching any of them |
contains | phone, email or link |
flag, min_severity | profanity, scam_accusation, sarcasm, human_review, emoji_only, mixed_language; severity 1 to 5 |
moderated_by | Why it was hidden or deleted: banned_words, rules, tone_filter, spam_filter, hide_all, legal_filter, blocked_author, you, with_parent, platform, other |
ai_reply, agent | sent, failed, none, edited; an agent id from /agents |
rule_task | open: comments your rules handed to you that are still open |
limit, cursor | 1 to 200, 100 by default |
A row
Every comment or message has id, type, platform, platform_id, created_at, updated_at, from_page, text, parent_id, channel, author, and post for comments or recipient_id for messages. Then:
analysis:language,sentiment,urgency,intent,is_question,severity,flags.hidden,deleted, andmoderationwhen either is true: by what (one of themoderated_bynames), the label and reason as your Inbox words them, the rule that fired, and when.rule_taskwhen one of your rules handed the comment to you:revieworreply_needed, when it was reviewed, the rule, and what it would have done.ai_reply:text,agent,published_at,response_seconds,delivered,failed,deleted,style.manual_reply: a reply written by a person.
post.text is the first 280 characters, and often empty on Instagram and TikTok; post.url is always there.
The other lists
/postsranks your posts and ads bycomments,negativeorrecentfor a period (limitup to 100, 25 by default; no cursor). Each comes with counts: comments, negative, positive, questions, replied, hidden, deleted./replieslists your agent's replies (sentby default, orfailed,edited), each with the customer's text it answered and how many seconds it took./summarycounts rows for a period, the last 7 days by default: total, comments, messages, positive, neutral, negative, urgent, questions, unanswered questions, replied, AI replies, hidden, deleted.group_bysplits them byday(up to 92 days, every day listed),channel,platform,sentiment,intentorlanguage./attentionreturns the conversations in Needs attention, in your Inbox's order, each with its reasons:needs_review,reply_needed,negative,urgent./channels,/agentsand/rulesreturn your setup as the dashboard shows it. They are read-only.
Actions
hide and unhide act on Facebook, Instagram or TikTok, like the buttons in your Inbox; review closes a comment your rules handed to you and changes nothing on the platform. They work only on customers' comments, on the channels the connection covers, with Moderate comments on.
With Change moderation rules on, POST /rules/cards adds a rule (text up to 300 characters, action one of keep, review, reply_needed, hide), PATCH /rules/cards/{id} edits one or switches it off, and PATCH /rules sets mode (review, apply, paused), default_action and banned_words (add, remove, action none or hide). Each answers your rules as they are after the change. Nothing can be set to delete: see Can a connected app change my moderation rules?
There is no endpoint to reply or delete.
MCP server
https://app.rypl.is/api/mcp, Streamable HTTP, with the same key in Authorization: Bearer. An app sees only the tools its connection allows:
| Tool | What it does |
|---|---|
list_channels | Your connected channels |
list_comments, list_messages | Rows, with the filters above, 25 per call and up to 100 |
list_posts | Which post or ad draws the most comments or complaints |
list_replies | Your agent's replies |
get_summary | Counts for a period |
list_attention | What needs your attention |
list_agents, get_moderation_rules | Your agents and Global rules |
hide_comment, unhide_comment, mark_comment_reviewed | Moderation, if you allowed it |
add_moderation_rule, update_moderation_rule, update_moderation_settings | Changing your Global rules, if you allowed it |