Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Agent Continuity Curator

    2

    Maintain durable, lean, and consistent AI agent memory across sessions while preventing context bloat and data leaks.

    $29

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Agent Continuity Curator

    Example session with this skill installed

    Clean up my Hermes memory. Add that I prefer early returns and Vitest for testing, and remove any info about the old legacy-auth-v1 project. Please do a safety check too.

    • Read your context and instructions
    • Compiled the agent continuity curator

    Updated ~/.hermes/memories/USER.md (1,120/1,375 chars)
    [+] Added Preference: User prefers early returns over nested if-blocks.
    [+] Added Fact: Uses Vitest for frontend testing.
    [-] Pruned: Obsolete project 'legacy-auth-v1' details.
    Privacy Check: No API keys or credentials detected.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Persistent Intelligence Without the Bloat

    The Agent Continuity Curator solves the short-lived session memory problem in AI agents. Instead of letting context windows fill up with redundant chat history and stale project details, this skill manages a structured system of durable memory files (like SOUL.md, MEMORY.md, and WORKING_MEMORY.md) to keep your agent consistent across sessions.

    What it does

    • Memory Architecture: Builds and maintains specialized files for persona (SOUL), user preferences (USER), and project state (WORKING_MEMORY).
    • Context Compression: Summarizes long chat histories into compact, actionable insights while pruning noise.
    • Hermes & Portable Modes: Supports both Hermes-native memory limits/delimiters and high-capacity portable formats for Claude, Cursor, and local agents.
    • Safety & Privacy: Audits memory content to exclude PII, tokens, secrets, and unnecessary personal data retention.
    • Intelligent Handoffs: Generates concise summaries that let a new agent instance pick up exactly where the last one left off.

    Why use this skill?

    Standard prompting often fails to maintain long-term behavioral consistency. This skill acts as a professional curator for your agent's working memory, enforcing strict memory budgets and organizational schemas. It helps your agent remember coding style and architectural decisions without wasting tokens on stale implementation details.

    • Cannot automatically detect non-textual secrets in binary files.
    • Memory pruning requires manual review if nuances are highly subjective.
    • Character limits are strictly enforced, potentially truncating long-form nuances.

    How to use

    1. 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. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 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 5 months ago

    Permissions

    Read Files
    Write Files

    File Scopes

    **/*.md

    Reads existing Markdown memory/handoff files and writes updated Markdown outputs such as MEMORY.md, USER.md, SOUL.md, USER_PROFILE.md, WORKING_MEMORY.md, OBSERVATIONS.md, and HANDOFF.md. No network, shell, browser, or environment-variable access is required.

    Listed5 months ago

    What's inside

    Frequently Asked Questions

    This skill manages an organized system of Markdown-based memory files to ensure your AI maintains a consistent persona and project awareness across different sessions without exceeding context limits.

    The skill is designed for versatility, supporting Hermes-native memory limits as well as high-capacity platforms like Claude, Cursor, and various local LLM environments.

    Unlike basic histories, the Curator actively prunes noise and compresses chat logs into structured summaries, ensuring only high-value information occupies your agent's context window.

    Safety is a priority; the skill includes auditing protocols that automatically identify and exclude PII, API tokens, and secrets from being stored in the durable memory files.

    You will receive the complete memory architecture framework, including the specialized schemas for SOUL, USER, and WORKING_MEMORY files, along with the logic for compression and handoffs.

    The system works by creating a 'handoff' summary at the end of a session, which acts as a bridge for the next instance to understand the current state and objectives immediately.

    New in Agents & Orchestration