Xainvault
Server Details
Scan AI skill packages, apply policy and human approval, and distribute approved versions via MCP.
- Status
- Healthy
- Uptime
- 100.0% over the last 24 hours
- Last Tested
- Transport
- Streamable HTTP · MCP 2025-11-25
- URL
- ChatGPT
- –
- Claude
- –
TDQS
Scored across 11 tools
Most tools map cleanly to distinct resources and actions (get_skill vs get_collection vs get_security_report vs check_skill_updates). The main overlap is list_my_skills vs search_skills, both of which return granted Ready skills and both expose a draft mode, so an agent must read descriptions to pick correctly. get_install_instructions vs create_install_session also sound similar but differ in that one returns setup docs and the other a signed download.
All 11 tools use snake_case verb_noun phrasing with a small, predictable verb set (list_, get_, create_, check_, save_, search_). The 'my_' qualifier in list_my_skills/list_my_groups is applied consistently and reads as personal-scope listing rather than an inconsistency.
Eleven tools is well within the sweet spot and each one covers a distinct slice of the vault workflow (discovery, retrieval, install, drafts, access, security). Nothing appears redundant enough to cut, and the surface is not so small that agents would hit dead ends.
Consumption side is solid: list/search, get, collection inspection, install session, install instructions, update checks and security reports. The authoring side dead-ends — save_skill_draft explicitly never submits, approves or releases, and there is no delete/update-draft or submit-for-review tool, so an agent cannot complete the draft lifecycle. Access tooling is also read-only (list_my_groups, no request/revoke).
Available Tools
11 toolscheck_skill_updatesCheck skill updatesCRead-onlyIdempotentInspect
Compare installed skill versions with active versions in the vault.
Inputs: installed (required). Returns: data, view, workspace, generated_at, +1 more
create_install_sessionCreate install sessionAIdempotentInspect
Create a 15-minute signed download for one granted Ready skill or collection. Select target=my_draft with skill_slug to test only the authenticated author's unreviewed Draft.
Inputs: target, skill_slug, collection_slug. Returns: data, view, workspace, generated_at, +1 more
get_collectionGet collectionBRead-onlyIdempotentInspect
Return a collection and the access/install state of each skill in it.
Inputs: slug (required). Returns: data, view, workspace, generated_at, +1 more
get_install_instructionsGet install instructionsBRead-onlyIdempotentInspect
Return direct MCP setup instructions for a supported developer client.
Inputs: client. Returns: data, view, workspace, generated_at, +1 more
get_security_reportGet security reportBRead-onlyIdempotentInspect
Return a trusted package security report visible to this connection.
Inputs: security_report_id (required). Returns: data, view, workspace, generated_at, +1 more
get_skillGet skillARead-onlyIdempotentInspect
Return a granted Ready skill by default, or select target=my_draft to inspect only this user's unreviewed Draft package.
Inputs: slug (required), target. Returns: data, view, workspace, generated_at, +1 more
list_collectionsList collectionsBRead-onlyIdempotentInspect
List collections visible in the active vault.
Inputs: limit, query, cursor. Returns: data, view, workspace, generated_at, +1 more
list_my_groupsList my groupsARead-onlyIdempotentInspect
List groups that grant access to the signed-in user in the active vault.
Inputs: limit, query, cursor. Returns: data, view, workspace, generated_at, +1 more
list_my_skillsList my skillsARead-onlyIdempotentInspect
List Ready skills granted to this user, or select state=draft to list only this user's authored Drafts.
Inputs: limit, query, state, cursor. Returns: data, view, workspace, generated_at, +1 more
save_skill_draftSave skill draftBDestructiveIdempotentInspect
Save a complete AI-authored skill package as a Draft in the active vault. Creates a new skill or a new version; never scans, submits, approves or releases it.
Inputs: mode (required), files (required), idempotency_key (required), expected_draft_id, +2 more. Returns: data, view, workspace, generated_at, +1 more
search_skillsSearch skillsARead-onlyIdempotentInspect
Search Ready skills by default, or select state=draft to search only this user's authored Drafts. Ready grants and requests do not apply to Drafts.
Inputs: limit, query, state, cursor. Returns: data, view, workspace, generated_at, +1 more
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
Scan AI agent skills, MCP server packages and MCP configs before you install them. Static analysis.
Governed AI agent skills — one library, distributed to devs and exposed to remote agents over MCP.
Scan GitHub-hosted AI skills for vulnerabilities: prompt injection, malware, OWASP LLM Top 10.
Discover and inspect public Runx skills before installing them. This read-only hosted MCP exposes search_skills and get_skill. The Apache-2.0 Rust runtime executes skills locally with scoped authority, approval gates and signed receipts. Docs: https://runx.ai/docs/mcp.
Related MCP Servers
- AlicenseAqualityBmaintenanceSecurity scanner for third-party AI agent-skill files: SKILL.md manifests, hooks, and bundled scripts, exposed via an MCP tool.186 npm3Apache 2.0

AgentAuditofficial
AlicenseAqualityDmaintenanceEnables AI agents to scan MCP servers and AI packages for vulnerabilities, prompt injection, and supply chain attacks.788 npmAGPL 3.0- AlicenseNot gradedqualityCmaintenanceEnables AI agents to search, retrieve, score, compose, and verify reusable agent skills from the SkillForge registry via MCP.21 npmMIT
- AlicenseNot gradedqualityDmaintenanceSecurity scanning for AI agent skills, MCP servers, and agent prompts, returning signed trust scores and detailed findings.MIT
Glama MCP Gateway
Add one secure layer between your agents and this server.