More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Enterprise RAG and Governance

    1

    Connect sources, control access, keep indexes fresh and cite sources. 10 ordered skills, starting with Enterprise Semantic Search, Internal Knowledge Assistants, RAG Access Control.

    Free

    Security scanned
    Enterprise RAG and Governance

    Enterprise RAG and Governance

    Example session with this skill installed

    Design an access control schema for a vector index that aggregates data from SharePoint and Jira, ensuring users only see documents they have permission to access.

    • Read your context and instructions
    • Compiled the enterprise rag

    I have designed a RAG access control schema that synchronizes source permissions into your vector database to enable query-time filtering.

    Enterprise RAG Access Control Schema

    Field NameTypeDescriptionPurpose
    doc_idStringImmutable source ID (e.g., SharePoint GUID)Stable reference for updates
    acl_metadataArrayList of User IDs, Group IDs, or RolesFilter target for query-time enforcement
    visibility_levelEnumpublic, internal, restricted, confidentialSecondary filter for safety overrides
    tenant_idStringUnique ID for the organization or departmentHard partitioning in multi-tenant setups
    last_synced_atTimestampISO 8601 date of last permission syncAudit for permission-sync lag
    source_systemStringsharepoint, jira, confluenceNamespace to prevent ID collisions

    Query-Time Enforcement Logic

    To prevent leakage, every retrieval request must include a filter block. For a user with ID u_123 and groups g_456, g_789:

    {
      "vector": [0.12, 0.45, ...],
      "filter": {
        "and": [
          { "tenant_id": "org_alpha" },
          {
            "or": [
              { "acl_metadata": { "contains": "u_123" } },
              { "acl_metadata": { "contains_any": ["g_456", "g_789"] } }
            ]
          }
        ]
      },
      "top_k": 5
    }
    

    Next steps

    1. Map your Identity Provider (Okta/AD) group IDs to the corresponding IDs in your SharePoint and Jira metadata.
    2. Implement a "deny-by-default" middleware in your retrieval service that injects these filters before hitting the vector store.
    3. Run an adversarial test by querying with a user ID known to lack access to a specific restricted document.

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

    About this skill

    The problem

    Companies deploying internal knowledge assistants often need a repeatable process, not a one-off prompt. This pack targets: Connect sources, control access, keep indexes fresh and cite sources.

    What's inside (10 skills, suggested learning order)

    1. Enterprise Semantic Search — Deploy meaning-based search on intranets.
    2. Internal Knowledge Assistants — Chatbots over company documentation and wikis.
    3. RAG Access Control — Apply per-user permissions to retrieval.
    4. Incremental Index Updates — Keep vector stores in sync.
    5. Citations and Source Traceability — Attribute answers to verifiable passages.
    6. Enterprise Data Connectors — Integrate SharePoint, Drive, Slack and CRMs with LLMs.
    7. Knowledge Base Curation — Structure and maintain source content for AI.
    8. Technical Documentation for AI — Write documentation optimized for LLM consumption.
    9. Email, Chat and Ticket Indexing for Search — Make internal conversations searchable by meaning.
    10. Conversation Memory with Retrieval — Combine chat history and RAG.

    Why a skill pack instead of a one-off prompt

    Each skill gives the agent a defined trigger, step-by-step procedure and output format, so results are consistent across sessions and easier to review.

    How to use

    1. Install the pack in your SKILL.md-compatible agent.
    2. Ask for the task in plain language.
    3. Review the output against the skill's quality checks before using it.

    Known limitations

    • Outputs are drafts that need human review; they are not professional, legal or financial advice.
    • Tools, APIs and prices change; check current vendor documentation before relying on version-specific details.
    • Quality depends on the context you give the agent (data, requirements, constraints).
    • Tested only where stated in the Compatibility note.

    Also by Rodrigo Santos

    Specialized Agents and Distribution; AI Agents: Fundamentals and Protocols; Model Fine-Tuning and Alignment

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for LLM Ops.

    Focuses on design and logic. Infrastructure, embedding models, and legal/HR approvals for data access must be provided by the user.

    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 buying a skill.

    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

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 3 days ago

    • Free with an Agensi account

    Software
    Specific MCP

    Permissions

    Read Files
    Listed3 days ago

    What's inside

    Frequently Asked Questions

    This skill provides architectural blueprints and logic for building 'Production RAG,' specifically addressing enterprise challenges like permission-aware retrieval, incremental data syncing, and verifiable source citations that simple prompts cannot handle.

    The package is framework-agnostic and compatible with major vector databases like Pinecone or Weaviate, as well as orchestration frameworks like LangChain, LlamaIndex, and implementations using the Model Context Protocol (MCP).

    Yes, a core feature of this skill is mapping complex source permissions (from platforms like SharePoint or Slack) into your vector index to ensure users only retrieve data they are authorized to see.

    You will receive a comprehensive implementation package including curation frameworks, metadata schemas, citation contract definitions, and validation test plans for security and retrieval accuracy.

    It utilizes change-detection patterns such as hashes and CDC (Change Data Capture) cursors to update your index incrementally, avoiding the high costs and downtime of full re-indexing.

    The skill focuses on the architectural logic and security layers; it does not include the physical infrastructure, specific embedding models, or SaaS API keys required to run the system.

    New in LLM Ops

    Free