canonical: https://jentic.com/apis/redreplier.com/redreplier

# RedReplier Public API

The RedReplier Public API monitors Reddit, X (Twitter), Bluesky, Hacker News, and Facebook for conversations that match a product's keywords, then scores each mention for relevance and drafts a suggested reply. You register the websites to monitor, manage each site's keyword set, and pull scored mentions filtered by platform, relevance band, minimum score, review status, and time window. Each mention can be marked approved or rejected, and an explain call fills in the relevance reason, tags, and AI reply suggestion when they are missing. Email alert cadence and keyword plan pricing are also available through the API.

## For AI agents

Pull keyword-matched Reddit, X, Bluesky, Hacker News, and Facebook mentions with relevance scores and drafted replies, approve or reject them, and manage the monitored sites and keywords.

## Scope

Does not post replies, send direct messages, or schedule social content. Use for monitoring and triaging keyword-matched social mentions only.

## Capabilities

- Pull scored mentions filtered by sources such as REDDIT_POST, HACKERNEWS, or BLUESKY, a minScore, and a from and to window
- Approve or reject a mention by setting its status to APPROVED or REJECTED
- Ask for the relevance reason, tags, and aiReplySuggestion behind a mention's score
- Register a website to monitor, letting RedReplier draft its product description from the URL
- Add, edit, disable, or re-enable keywords for a monitored website
- Price a keyword count or the pending keywords before activating them, without being charged
- Turn email alerts on or off and set their cadence in minutes

## Use cases

### Mention triage agent through Jentic

A founder wants an agent to sift social conversations each morning and surface the few worth answering. Through Jentic the agent searches for 'list new RedReplier mentions', loads the mentions operation with its input schema, and pulls NEW mentions with a minScore and HIGH or VERY_HIGH score buckets while the bearer token is injected at call time. It reads each mention's drafted reply, sets the status to APPROVED or REJECTED, and leaves posting to the human.

Example prompt: Call GET /mentions with statuses NEW, scoreBuckets HIGH and VERY_HIGH, sort RELEVANCE, and limit 20, then PATCH `/mentions/{id}/status` with APPROVED or REJECTED for each

### Keyword set tuning

Marketers refine which phrases RedReplier watches as they learn what produces useful threads. The API adds keywords to a website, where those within the plan allowance become ACTIVE and the rest PENDING, edits a keyword's text without losing its paid slot, and disables a keyword to stop monitoring while keeping its mentions. Deleting a keyword is permanent and removes its mentions.

Example prompt: Call POST `/websites/{id}/keywords` with keywords ['invoice automation', 'ap software'], then POST `/keywords/{id}/disable` for a keyword that only returns noise

### Weekly mention volume report

Teams tracking community interest want counts, not rows. The count operation accepts the same filters as the mentions list, including sources, keywords, statuses, and a from and to window, so a report can show how many relevant conversations each platform produced in a week without downloading every mention.

Example prompt: Call GET `/mentions/count` once per source, REDDIT_POST, TWITTER, BLUESKY, and HACKERNEWS, with from set to seven days ago, and tabulate the results

### Onboard a new product for monitoring

When a new product launches, it needs its own monitored website so mentions are scored against what it does. The analyze-description operation drafts a description from the product URL without creating anything, and creating the website takes a url plus optional name, keywords, and description; leaving the description out lets RedReplier write one, which uses an AI generation from the plan quota.

Example prompt: Call POST `/websites/analyze-description` with the product url, review the returned description, then POST /websites with url, name, description, and an initial keywords list

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| GET | `/mentions` | List scored mentions |
| PATCH | `/mentions/{id}/status` | Approve or reject a mention |
| POST | `/mentions/{id}/explain` | Explain a mention's score |
| GET | `/mentions/count` | Count matched mentions |
| POST | `/websites` | Start monitoring a website |
| POST | `/websites/{id}/keywords` | Add keywords to a website |
| POST | `/keywords/{id}/disable` | Stop monitoring a keyword |
| PUT | `/alert-settings` | Set email alert cadence |

## Key resources

- **Mentions** — Keyword-matched conversations with relevance scores, reasons, tags, AI reply drafts, and review status
- **Websites** — Monitored products with their descriptions and keyword sets
- **Keywords** — Monitored phrases that can be added, edited, disabled, enabled, or activated from pending
- **Billing previews** — Plan pricing for a keyword count or for activating pending keywords, and the keyword-change allowance
- **Alert settings** — Email alert on or off and cadence in minutes, clamped to the plan floor
- **Workspaces** — Workspaces the caller can act in, selected with the X-Workspace-Id header

## AI readiness

This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use.

- **Score:** 67 / 100
- **Maturity:** AI-Aware
- **Dimensions:**
  - Foundational Compliance: 99 / 100
  - Developer Experience & Jentic Compatibility: 63 / 100
  - AI-Readiness & Agent Experience: 49 / 100
  - Agent Usability: 94 / 100
  - Security: 60 / 100
  - AI Discoverability: 71 / 100
- **View full report:** https://jentic.com/apis/redreplier.com/redreplier/scorecard
- **How the score is calculated:** https://docs.jentic.com/reference/api-readiness-framework/overview/
- **More about the dimensions:** https://docs.jentic.com/reference/api-readiness-framework/specification/#dimensional-model-overview

### Score it yourself

Every API in the directory is allowlisted, so you can re-score it with no key required.

- **Score your own API:** https://jentic.com/scorecard.md
- **Scoring CLI agent skill:** https://github.com/jentic/jentic-api-scorecard/blob/main/skills/jentic-api-scorecard/SKILL.md

```sh
npx @jentic/api-scorecard-cli score <openapi-url>
```

## Why Jentic

- **Setup:** Wiring RedReplier by hand means issuing a bearer token, handling the X-Workspace-Id header when a sign-in spans several workspaces, and pacing calls from the RateLimit headers. Through Jentic you install once, import RedReplier from the API Directory, store the token once, and your agent calls it.
- **Permission scoping:** RedReplier puts the mention, keyword, and website ids in the URL path (`/mentions/{id}/status`, `/keywords/{id}`), so a rule can bound your agent to particular records and the operations you pick. You choose the operations it may call, so destructive ones like deleting a keyword with its mentions or removing a website are not included unless you add them.
- **Credential handling:** Your RedReplier token is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'list new RedReplier mentions' or 'approve a mention', and Jentic returns the matching RedReplier operation with its input schema, including the sources, scoreBuckets, and statuses enums.

## Related APIs

- **Mention API** — Brand monitoring across web and social sources
- **AIMentionTracker API** — Tracks how brands are mentioned in AI assistant answers
- **Reddit API** — Read and post to Reddit threads
- **Hacker News API** — Items, comments, and users on Hacker News

## FAQ

### What authentication does the RedReplier Public API use?

Per its OpenAPI spec, the RedReplier Public API uses HTTP bearer authentication: send an API token from the API Tokens page of the dashboard as Authorization: Bearer. The spec notes that an OAuth access token can reach several workspaces, chosen per call with the X-Workspace-Id header, while an API key reaches only its own. Through Jentic the token is stored by your own instance and attached at call time.

### Does the RedReplier API post replies to Reddit for me?

No. The API returns an AI-drafted reply with each mention and lets you mark the mention APPROVED or REJECTED with PATCH `/mentions/{id}/status`, but this spec has no operation that posts to Reddit, X, Bluesky, Hacker News, or Facebook. Posting stays with you.

### What are the rate limits for the RedReplier Public API?

The spec does not state a numeric limit, but it says limits are counted per API key and every response carries RateLimit-Policy, RateLimit-Limit, RateLimit-Remaining, and RateLimit-Reset headers, with Retry-After on a 429. Read those headers to pace calls, and check https://redreplier.com/ for plan details.

### Is there a RedReplier MCP server?

You don't need an MCP server to give your agent RedReplier. Jentic connects it directly from the API Directory: import it, store your API token once, and your agent calls the mention, keyword, and website operations on demand without loading extra tool definitions into its context.

### How do I pull today's best mentions through Jentic?

Search Jentic for 'list new RedReplier mentions', load GET /mentions, and run it with statuses NEW, a from timestamp for the start of today, and sort RELEVANCE. Each result includes the matched keyword, its score, and a drafted reply. To run it on your own infrastructure, install Jentic One from its GitHub repo.

### Will changing keywords through the API charge my card?

The spec says enabling a keyword and activating pending keywords never charge: keywords beyond the plan stay PENDING until you upgrade in the RedReplier app. GET `/keywords/billing-preview` and GET `/keywords/activate-pending/preview` show what an upgrade would cost without charging anything.

### Can I limit what my agent is allowed to do with the RedReplier Public API?

Yes. For example, allow only GET /mentions and PATCH `/mentions/{id}/status`, so the agent can read and review mentions but cannot delete keywords, remove websites, or change alert settings, and every call it makes is logged.