Works with the AI Agents you already use
saintstack memory
Persistent cross-session memory and architectural decision tracking for AI coding agents via SaintStack MCP.
$2
saintstack memory
Example session with this skill installed
I'm starting a new session on the 'saint-api' project. Get the current context and check if we've already decided on a validation library for the auth module.
- Read your context and instructions
- Compiled the saintstack memory
I have retrieved the project context and searched the memory for architectural decisions regarding validation libraries. ## Project Context & Memory Audit | Category | Details | | :--- | :--- | | **Project ID** | `saint-api` | | **Last Session** | Refactored middleware and added JWT verification logic. | | **Status** | 2 tasks completed; `src/auth/validate.ts` created but pending implementation. | | **Active ADR** | **ADR-004: Zod for Schema Validation** | | **Decision** | Use `zod` for all request body and environment variable validation. | | **Rationale** | Strong TypeScript inference and compatibility with existing Hono middleware. | | **Alternatives** | `Joi` (rejected due to weaker TS support) and `Valibot` (rejected for ecosystem maturity). | ### Search Results: "validation library" 1. **Decision Found:** Zod was accepted on 2024-05-12 for the `auth` and `billing` modules. 2. **Implementation Note:** Ensure all schemas are exported from `@/schemas` to maintain consistency. 3. **Open Issue:** `auth` module needs a Zod schema for the login endpoint. **Next steps** - Initialize `zod` in the `auth` module as per ADR-004. - Create the login request schema in `src/auth/schemas.ts`. - Update the session log once the validation logic is linked to the middleware.
Connects securely to your tools. The creator never sees your data.
About this skill
Large codebases exhaust context windows, forcing agents to re-scan files and lose track of architectural decisions. This skill provides persistent, cross-session memory for AI coding agents via the SaintStack MCP gateway. It ensures that technical choices, session progress, and project invariants survive beyond a single chat window. ## What it does - **Persistent Context Retrieval** queries the SaintStack gateway for project-wide architecture and rules instead of re-scanning the file tree. - **Architectural Decision Recording** logs technical choices, rationales, and rejected alternatives as structured ADRs. - **Session Handover Logging** captures completed tasks, modified files, and open issues to bridge gaps between separate agent runs. - **Code Graph Ingestion** stores high-level module maps and dependency relations from tools like Graphify. - **Semantic Memory Search** locates specific coding patterns or conventions within previous project history. ## How it works 1. **Initialize** by calling `saintstack_get_context` at the start of a task to load project invariants and the last session state. 2. **Search** via `saintstack_search_memory` when looking for specific technical patterns or legacy decisions. 3. **Commit** technical choices using `saintstack_save_decision` to prevent architectural drift. 4. **Finalize** every session with `saintstack_save_session` to prepare the context bundle for the next agent or device. ## Frameworks & tools This skill requires the SaintStack MCP gateway. It is designed for use with Claude Code, Cursor, and other MCP-compatible AI coding environments. ## Why this beats prompting it yourself Standard prompts rely on the local file system or chat history, both of which are ephemeral or token-heavy. This skill offloads project knowledge to a persistent layer, reducing token waste and ensuring architectural consistency across different agents and devices. ## Use cases - Resuming a complex refactor across different IDE sessions or devices. - Onboarding a new agent to a codebase without a full file-by-file scan. - Maintaining a living ADR log that the agent consults before adding dependencies. - Synchronizing project state between desktop AI and mobile coding interfaces. ## Known limitations Requires an active SaintStack MCP gateway connection. Memory retrieval is scoped to specific project IDs and does not include raw secrets or environment variables.
Requires an active SaintStack MCP gateway connection. Does not store API secrets or private credentials. Project ID must be manually identified or provided.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 6 days ago
Requires
Permissions
Allowed Hosts