Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Mergify 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%2Fmergify.com%2Fmergify" | 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%2Fmergify.com%2Fmergify" | 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 Mergify API.
Read the current merge queue status for a branch and inspect a single queued pull request by pr_number
Preview the dequeue impact of a pull request, then remove it from the merge queue
Pause the merge queue with a required reason and resume it when the incident is over
Schedule a merge freeze with reason, start, end, and timezone, then update or delete it
GET STARTED
Search CI tests in a repository and read a test's failures, executions, and health transitions
Quarantine a flaky test by test_name and reason, optionally with unquarantine_when_healthy
Pull merge queue statistics such as queue size, failure rate, batch size, and average CI runtime
Patterns agents use Mergify API for, with concrete tasks.
★ Agent-driven merge queue triage through Jentic
A coding agent watching a GitHub repository finds that merges have stalled. Through Jentic it calls the Mergify merge-queue status operation for the affected branch, reads the queued pull request that is failing, previews the dequeue impact, and removes it so the rest of the queue can proceed. The Mergify token stays on your own Jentic One instance, and your rules decide whether dequeue is one of the operations the agent may call.
Call GET /repos/acme/web/merge-queue/status with branch 'main', pick the failing pull request, call GET /repos/acme/web/merge-queue/pull/412/dequeue-impact, then POST /repos/acme/web/merge-queue/pull/412/dequeue
Release and holiday merge freezes
Teams that stop merging during a release window or holiday can create a scheduled freeze instead of asking everyone to hold off. The create operation requires a reason, a start time, and a timezone, with an optional end and matching_conditions or exclude_conditions to let hotfixes through. Freezes can be listed, updated by freeze_id, and deleted when plans change.
POST /repos/acme/web/scheduled_freeze with reason 'Q4 release', start '2026-12-20T00:00:00', end '2026-12-27T00:00:00', timezone 'Europe/Dublin'
Flaky test quarantine
Flaky tests make merge queue runs fail for reasons unrelated to the change. With CI Insights an agent searches tests in a repository, reads a suspect test's failures and health transitions, and marks it quarantined with test_name and reason. Setting unquarantine_when_healthy lets the test come back once it stabilises, and the quarantine can be removed by quarantine_id.
Call GET /ci/acme/repositories/web/search/tests with a test_name filter, read GET /ci/acme/repositories/web/tests/{test_id}/failures, then POST /ci/acme/repositories/web/quarantines with a test_name filter and reason 'flaky on retry'
Merge queue health reporting
Engineering leads tracking delivery speed can pull merge queue statistics per repository: queue size, failure rate, batch size, checks retries, idle and total queue time, and the CI runs the queue started and wasted. Each statistic accepts base_ref and queue_name filters so a report can isolate one branch or queue.
Call GET /repos/acme/web/stats/failure_rate and GET /repos/acme/web/stats/average_ci_runtime with base_ref 'main' and summarise both for the weekly report
65 endpoints — the mergify api controls the mergify merge queue for github repositories and reports on the ci runs behind it.
METHOD
PATH
DESCRIPTION
/repos/{owner}/{repository}/merge-queue/status
Get the current merge queue status
/repos/{owner}/{repository}/merge-queue/pull/{pr_number}/dequeue
Dequeue a pull request
/repos/{owner}/{repository}/merge-queue/pause
Pause the merge queue
/repos/{owner}/{repository}/scheduled_freeze
Create a scheduled freeze
/ci/{owner}/repositories/{repository}/search/tests
Search CI tests
/ci/{owner}/repositories/{repository}/quarantines
Quarantine a test
/repos/{owner}/{repository}/stats/failure_rate
Get the queue failure rate
/repos/{owner}/{repository}/merge-queue/status
Get the current merge queue status
/repos/{owner}/{repository}/merge-queue/pull/{pr_number}/dequeue
Dequeue a pull request
/repos/{owner}/{repository}/merge-queue/pause
Pause the merge queue
/repos/{owner}/{repository}/scheduled_freeze
Create a scheduled freeze
/ci/{owner}/repositories/{repository}/search/tests
Search CI tests
/ci/{owner}/repositories/{repository}/quarantines
Quarantine a test
/repos/{owner}/{repository}/stats/failure_rate
Get the queue failure rate
What agents get from Jentic-routed access to this vendor.
Setup
Wiring Mergify by hand means choosing between four bearer credentials (admin key, ci key, Mergify user token, GitHub token) per operation and tracking which queue endpoints are deprecated. Through Jentic you install once, import Mergify from the API Directory, store the token once, and your agent calls it.
Permission scoping
Mergify puts the owner and repository in the URL path (/repos/{owner}/{repository}/...), so a rule can pin your agent to one repository and the operations you pick. You choose the operations it may call, so actions like dequeue, queue pause, or freeze deletion are not included unless you add them.
Credential isolation
Your Mergify 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.
Intent-based discovery
Agents search Jentic by intent such as 'check the merge queue status' or 'quarantine a flaky test', and Jentic returns the matching Mergify 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 Mergify API through Jentic.
What authentication does the Mergify API use?
Per its OpenAPI spec, the Mergify API uses bearer tokens in the Authorization header, with four declared schemes: a Mergify application key with the admin scope (AdminApplicationKey), an application key with the ci scope (CIApplicationKey), a Mergify user token (MergifyTokenBearerAuth, mut_ prefix), and a GitHub user access token (GitHubTokenBearerAuth). Each operation lists which it accepts: most merge queue and CI Insights operations take an admin key, user token, or GitHub token, while the scope-setting operations take the ci key. Through Jentic the token is stored encrypted by your own Jentic One instance and injected when the agent calls.
Can I quarantine a flaky test with the Mergify API?
Yes. POST /ci/{owner}/repositories/{repository}/quarantines takes a required test_name and reason, plus optional branch and unquarantine_when_healthy. You can list quarantined tests, read one by quarantine_id, and remove it with DELETE on the same quarantine_id path.
How do I pause the merge queue through Jentic?
Install Jentic One with `curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh`, import the Mergify API from the directory, and store your Mergify token. Your agent then searches 'pause the Mergify merge queue', gets PUT /repos/{owner}/{repository}/merge-queue/pause with its schema, and sends the required reason. DELETE on the same path resumes the queue. To run it on your own infrastructure, install Jentic One from its GitHub repo.
Can I limit what my agent is allowed to do with the Mergify API?
Yes. For example, write a rule that allows GET /repos/acme/web/merge-queue/status and the CI Insights test reads under /ci/acme/repositories/web, so the agent can diagnose a stalled queue in that one repository but cannot dequeue, pause, or quarantine anything. Add the dequeue operation later if you want it to act, and every call it makes is logged.
Is there a Mergify MCP server?
You don't need an MCP server to give your agent Mergify. Jentic connects it directly from the API Directory: import it, store your credential once, and your agent calls it. Operations are discovered on demand, so no extra tool definitions sit in the agent's context.
What are the rate limits for the Mergify API?
The OpenAPI spec does not specify rate limits. Check the Mergify API documentation at https://docs.mergify.com/api/usage for current limits.
Which pause endpoint should I use, queue/pause or merge-queue/pause?
Use /repos/{owner}/{repository}/merge-queue/pause. The older /queue/pause operations are marked deprecated in the spec, as are several older queue and stats endpoints, so new integrations should target the merge-queue paths.
For Agents
Check and steer the Mergify merge queue for a GitHub repository: read queue status, dequeue or pause, schedule freezes, and quarantine flaky tests found through CI Insights.
Use for: Check whether the merge queue on main is stuck, Pause the Mergify merge queue while we fix a broken deploy, Find all flaky tests in this repository and quarantine the worst one, Get the failure history for a specific test
Not supported: Does not host repositories, run CI jobs, or review code. Use for controlling the Mergify merge queue and CI Insights test data only.
The Mergify API controls the Mergify merge queue for GitHub repositories and reports on the CI runs behind it. It reads the live queue status for a branch, dequeues a pull request after previewing the impact, pauses or resumes the queue, and schedules merge freezes with a start, end, and timezone. It also exposes CI Insights: search tests, read a test's failures, executions, and health transitions, and quarantine flaky tests so they stop blocking merges.
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>