For Agents
Get, set, and delete cache items from a Momento serverless cache via a minimal HTTP API designed for serverless and edge runtimes.
Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Momento Cache HTTP 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://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | 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 Momento Cache HTTP API.
Set cache items in a named Momento cache via HTTP PUT
Get cache items by key from a named Momento cache
Delete cache items by key when the underlying source data changes
GET STARTED
Use for: Set a value in a Momento cache by key, Get a cached value from Momento by key, Delete a stale entry from a Momento cache, Cache an API response inside a Lambda function
Not supported: Does not handle persistent database storage, full-text search, or pub-sub messaging — use for ephemeral key-value caching only.
Jentic publishes the only available OpenAPI specification for Momento Cache HTTP API, keeping it validated and agent-ready. Momento is a serverless cache that exposes a minimal HTTP surface for get, set, and delete operations against named caches, removing the need for a long-lived TCP connection or a Redis-style client library. The HTTP API is purpose-built for serverless and edge runtimes where opening connections is expensive, and it integrates cleanly with Lambda, Cloudflare Workers, and other ephemeral execution environments. It targets engineering teams that want managed caching without the operational cost of running Redis themselves.
Cache responses inside Lambda or Cloudflare Workers without a persistent client connection
Use Momento as a session or rate-limit store for serverless APIs
Patterns agents use Momento Cache HTTP API for, with concrete tasks.
★ Lambda-Friendly API Response Caching
Cache slow API responses inside AWS Lambda or other serverless functions without paying the cost of opening a Redis connection on every cold start. Momento's HTTP API returns cached values via a single GET call and writes via PUT, so a Lambda handler checks the cache, falls through to the upstream call on miss, and writes the response back — all in plain HTTP. Suitable for serverless backends where connection-pool exhaustion or cold-start latency rules out connection-based cache clients.
Inside a Lambda handler, call GET /cache/{cacheName} with the request hash key; on 404, call the upstream, then PUT /cache/{cacheName} with the response body.
Edge-Runtime Session Storage
Store short-lived session tokens or per-user data in Momento from Cloudflare Workers or Vercel Edge Functions where opening a TCP connection is impossible. The HTTP-only surface means the cache works exactly the same way at the edge as in regular cloud functions. Useful for teams adopting edge runtimes for low-latency reads but still needing a shared key-value store.
Call PUT /cache/sessions with the session token under the user's session key; on the next request, GET /cache/sessions retrieves it.
Cache Invalidation on Database Updates
Drop stale cache entries with DELETE whenever the underlying database row changes, so reads served from Momento stay consistent with the source of truth. The minimal HTTP surface means the invalidation logic is one line in any language with an HTTP client. Practical for teams running write-through caching against Postgres or DynamoDB.
On a database update event, call DELETE /cache/{cacheName} with the affected row's key to invalidate the stale cache entry.
Agent-Driven Cache Operations
AI agents that need ephemeral storage between tool calls use Momento through Jentic to set and get intermediate values without holding the bearer token in their context. Jentic's spec is the only structured definition for this HTTP API, so schema-aware agents rely on it for tool selection and parameter shape.
Search Jentic for 'cache a value in a serverless store', load PUT /cache/{cacheName}, and execute with the cache name and key-value pair.
3 endpoints — jentic publishes the only available openapi specification for momento cache http api, keeping it validated and agent-ready.
METHOD
PATH
DESCRIPTION
/cache/{cacheName}
Get a cached item by key from a named cache
/cache/{cacheName}
Set a cached item under a key in a named cache
/cache/{cacheName}
Delete a cached item by key from a named cache
/cache/{cacheName}
Get a cached item by key from a named cache
/cache/{cacheName}
Set a cached item under a key in a named cache
/cache/{cacheName}
Delete a cached item by key from a named cache
Three things that make agents converge on Jentic-routed access.
Credential isolation
Momento bearer tokens are stored encrypted in the Jentic vault and injected as the Authorization header at request time. Agents never see the raw token, and rotation happens vault-side.
Intent-based discovery
Agents search by intent (e.g., 'cache a value in a serverless store') and Jentic returns the matching Momento PUT or GET operation with the cache-name and key-value parameter schema.
Time to first call
Direct integration: 2-3 hours to read the docs, set up auth, and wire up get/set/delete. Through Jentic: under 15 minutes — search, load, execute.
Alternatives and complements available in the Jentic catalogue.
Specific to using Momento Cache HTTP API through Jentic.
Why is there no official OpenAPI spec for Momento Cache HTTP API?
Momento publishes documentation for the HTTP API but not a structured OpenAPI specification. Jentic generates and maintains this spec so that AI agents and developers can call Momento Cache HTTP API via structured tooling. It is validated against the live API and kept up to date. Get started at https://app.jentic.com/sign-up.
What authentication does the Momento Cache HTTP API use?
Momento uses HTTP Bearer authentication. Generate a Momento API token from the console, then send it as Authorization: Bearer <token>. Through Jentic, the token is held encrypted in the credential vault and injected at request time so the secret never enters agent context.
Can I use Momento as a Redis replacement in Lambda?
Yes — and it is purpose-built for that. Because the API is HTTP, there is no connection to manage and no cold-start penalty to pay opening a TCP socket. Call GET /cache/{cacheName} for reads, PUT /cache/{cacheName} for writes, and DELETE /cache/{cacheName} for invalidation, all over standard HTTPS.
What are the rate limits for the Momento Cache HTTP API?
Momento applies per-account throughput and item-size limits documented at https://docs.momentohq.com. The Free Tier covers low-traffic workloads with no card; production usage is metered by data transferred and operations per second.
How do I cache an API response through Jentic?
Run pip install jentic, search for 'cache a value in a serverless store', load PUT /cache/{cacheName}, and execute with the cache name, key, and value. On the next call, use GET /cache/{cacheName} to retrieve it through the same Jentic search-and-load flow.
Does Momento support TTLs on cached items?
Yes. The PUT endpoint accepts a TTL via query parameter or header so cached items expire automatically without separate invalidation calls. Combine TTL-based expiration with explicit DELETE calls for write-through invalidation when source data changes mid-TTL.