Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Huddlekit API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fhuddlekit.com%2Fhuddlekit" | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fhuddlekit.com%2Fhuddlekit" | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
What an agent can do with Huddlekit API.
List every website project, web app, and document in the workspace to find the parent_id a comment belongs to
Pull the newest comments on one project, web app, or document, up to 200 per call
Post a comment on a page path, a web app URL, a document page number, or a video timestamp
Move a comment between open, in-review, in-progress, and resolved
GET STARTED
Subscribe an https URL to comment.created or comment.status_changed events and remove that subscription later
Render sample webhook payloads from recent comments to map fields before any event fires
Patterns agents use Huddlekit API for, with concrete tasks.
★ Agent triage of design review comments through Jentic
A developer's coding agent pulls the open review comments on a website project through Jentic, fixes what it can in the codebase, and moves each handled comment to in-review. The agent finds the operations by intent, the Huddlekit key is injected by the developer's own Jentic One instance, and a rule can keep the agent on read and status-change calls only.
Call GET /projects to find the website project named Marketing Site, call GET /comments with its id as parent_id, then for each comment fixed in the repo call PATCH /comments/{id} with status in-review
Feed new comments into an issue tracker
Send every new Huddlekit comment to a tracker or chat channel as it is created. Subscribing an https URL to comment.created events gives a REST-hook delivery signed with the X-Huddlekit-Signature header, and the sample payloads call shows the exact delivery shape so the receiving handler can be built and tested first.
Call GET /events/recent with event comment.created and limit 3 to get sample payloads, then call POST /hooks with targetUrl https://hooks.example.dev/huddlekit and event_types comment.created
Post QA findings as page comments
A QA script or agent that finds a layout bug can leave the finding where the team already reviews work. POST /comments takes the parent_id, the surface, the comment text, and a page path such as /pricing for a website, a url for a web app, or a page_number for a document.
Call POST /comments with the website project's parent_id, surface website, path /pricing, and text 'Primary button overlaps the footer at 375px width'
Weekly feedback status report
Summarise how much review feedback is still open across a workspace. The agent lists all projects, web apps, and documents, pulls the newest comments on each, and counts them by status so a team lead can see what is still open, in review, in progress, or resolved.
Call GET /projects, then GET /comments for each parent_id with limit 200, and report the count of comments in each status: open, in-review, in-progress, and resolved
9 endpoints — the huddlekit api reads and writes the review comments a workspace leaves on websites, web apps, and uploaded documents.
METHOD
PATH
DESCRIPTION
/projects
List projects, web apps, and documents
/comments
List comments on one parent
/comments
Create a comment
/comments/{id}
Change a comment's status
/hooks
Subscribe a URL to comment events
/hooks/{id}
Unsubscribe a webhook
/events/recent
Get sample webhook payloads
/me
Check the key's workspace and scopes
/projects
List projects, web apps, and documents
/comments
List comments on one parent
/comments
Create a comment
/comments/{id}
Change a comment's status
/hooks
Subscribe a URL to comment events
/hooks/{id}
Unsubscribe a webhook
/events/recent
Get sample webhook payloads
/me
Check the key's workspace and scopes
What agents get from Jentic-routed access to this vendor.
Setup
Wiring the Huddlekit API by hand means creating a workspace key with the right read or write scope, sending it as a bearer token, looking up each parent_id and surface before you can touch a comment, and backing off on 429 responses. Through Jentic you install once, import the Huddlekit API from the API Directory, store the key once, and your agent calls it.
Permission scoping
Huddlekit takes the project, web app, or document id as the parent_id query or body field rather than in the URL path, so rules bound which operations your agent may call, not which project. You choose the operations it may call, so posting comments or subscribing webhooks are not included unless you add them.
Credential isolation
Your Huddlekit workspace key 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.
Intent-based discovery
Agents search Jentic by intent such as 'list feedback comments' or 'resolve a comment', and Jentic returns the matching Huddlekit operation with its input schema so the agent calls the right endpoint without browsing the reference docs.
Alternatives and complements available in the Jentic catalogue.
Specific to using Huddlekit API through Jentic.
What authentication does the Huddlekit API use?
Per its OpenAPI spec, the Huddlekit API uses a workspace API key sent as a bearer token in the Authorization header, in the form hk_live_ followed by 64 hex characters. Keys carry a read scope for GET calls, a write scope for POST, PATCH, and DELETE calls, or both, and GET /me tells you which scopes a key has. Through Jentic the key is stored encrypted by your own Jentic One instance and never handed to the agent.
Can I edit or delete comments with the Huddlekit API?
No. The spec states that status is the only field PATCH /comments/{id} can change: sending text returns 400, and there is no delete operation for comments. You can create comments with POST /comments and move them between open, in-review, in-progress, and resolved.
What are the rate limits for the Huddlekit API?
Huddlekit's OpenAPI spec states that limits are counted per API key, per endpoint group, and separately for reads and writes: up to 200 reads and 30 writes a minute. Going over returns 429 with a Retry-After header in seconds, and refused calls count toward the limit too.
Does the Huddlekit API need a paid plan?
According to the spec, every call except DELETE /hooks/{id} requires an active or trialing Team subscription, and a past-due payment fails the check with a 403 that lists requiredPlans. Deleting a webhook is deliberately not plan-gated so automations can always be switched off.
How do I resolve a Huddlekit comment through Jentic?
Search Jentic for 'resolve a review comment' to get PATCH /comments/{id}, then call it with the comment id and status resolved, passing surface when the comment is on a web app or document. To run it on your own infrastructure, install Jentic One from its GitHub repo, add the Huddlekit API from the directory, and store your key once.
Is there a Huddlekit MCP server?
You don't need an MCP server to give your agent Huddlekit. Jentic connects it directly from the API Directory: import it, store your workspace key once, and your agent calls the operations you allow, without another server's tool definitions sitting in its context.
Can I limit what my agent is allowed to do with the Huddlekit API?
Yes. For example, write a rule that allows only GET /projects, GET /comments, and PATCH /comments/{id}, so the agent can read feedback and change comment statuses but cannot post new comments or create webhook subscriptions. Every call it makes is logged by your Jentic One instance.
For Agents
Read website, web app, and document review comments in a Huddlekit workspace, post new comments, change a comment's status to resolved or in-progress, and subscribe a URL to comment events.
Use for: List the open feedback comments on our marketing site project, Mark comment 3f2a as resolved, Post a comment on the /pricing page saying the CTA overlaps the footer, Find all comments left on the onboarding PDF
Not supported: Does not edit or delete comment text, create projects or upload documents, or manage workspace members and billing. Use for reading, posting, and triaging Huddlekit review comments and their webhooks only.
The Huddlekit API reads and writes the review comments a workspace leaves on websites, web apps, and uploaded documents. It lists a workspace's projects, web apps, and documents, returns the newest comments on any of them, adds new comments, and moves a comment through the open, in-review, in-progress, and resolved statuses. It also manages REST-hook webhook subscriptions for comment events, signed with HMAC-SHA256, and can render sample payloads from recent comments so you can map fields before a live event fires. Comment text cannot be edited and comments cannot be deleted through the API.
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.
Base layer of spec validity and structural soundness.
Aggregated quality score from linter diagnostics, weighted by severity.
Percentage of `$ref` references that resolve successfully.
Checks whether the API description parses successfully and conforms to its declared specification (e.g., OpenAPI).
Structural correctness score based on schema issues using logarithmic dampening.
Clarity, completeness, and ingestion readiness for developers and tooling.
How richly the API is illustrated with examples.
Percentage of examples that conform to their schemas.
Percentage of operations with complete response definitions (success, client error, server error).
Health of API ingestion, bundling, and resolution within Jentic pipelines.
Semantic breadth, depth, and agent comprehension for AI systems.
Coverage of descriptions across API elements.
Coverage of RFC 9457 Problem Details for error responses.
Coverage, uniqueness, and casing consistency of operationIds for AI inference.
Coverage of summaries across operations/tags/info.
Functional utility, complexity comfort, and AI orchestration readiness.
Agent comfort level based on API operational and structural complexity.
Trust, risk posture, and security compliance.
Average quality of security schemes based on authentication method strength (weakest link for OAuth2).
Findability, semantic richness, and reasoning readiness.
Clarity and depth of descriptions across API elements.
Score it yourself
Every API in the directory is allowlisted, so you can re-score it with no key required.
npx @jentic/api-scorecard-cli score <openapi-url>