Loading ToolBench

Reading the current corpus…

pocketbase-mcp-server MCP server · D · ToolBench

← Browse

pocketbase-mcp-server

github.com/imiborbas/pocketbase-mcp-server

Model Context Protocol Server for PocketBase

2026-07-28+STDIO1 tools

50/100

Definition Quality50

Protocol Readiness50

SupportabilityNot measured-

Spec posture

Inferred effective spec 2026-07-28+·current baseline 2026-07-28

Static source inference · medium confidence

Current-spec patterns detected

Summary

This MCP server has only one tool with a minimal definition. The tool has a valid name and basic schema, but lacks critical definition quality. The input schema is empty (no parameters), which is acceptable for a read-only list operation, but the description is generic and lacks context about when to use this tool or what it returns. No output schema is documented anywhere. The server exposes resources and prompts capabilities but provides no visible definitions for them in the source code provided. Overall, the definition quality is well below production baseline.

Tools (1)

pocketbase_list_collectionsread onlyauthsource verified50/100

List PocketBase Collections

Issues (5)

CRITICALtool-schema ↗

No output schema documented for pocketbase_list_collections. LLMs cannot infer what fields are returned (e.g., collection name, ID, type, fields). This violates the requirement that tool outputs must be documented so agents can plan downstream operations.

pocketbase_list_collections

HIGHtool-description ↗

Description is only 26 characters ('List PocketBase Collections'). It lacks context about WHEN to call this tool, what data structure is returned, or whether it's a discovery tool to precede other operations.

pocketbase_list_collections

MEDIUMconstrained-input ↗

Input schema is empty (no parameters, no required fields). While this is acceptable for a read-only list, the schema scores low because there is no documentation of what filters or options a user might need. A complete schema would include optional parameters like page_size, offset, or filter_by_name.

pocketbase_list_collections

Recommendations

Expand the description of pocketbase_list_collections to 80-150 characters. Example: 'Retrieve all collections from the PocketBase database. Call this first to discover available tables and their metadata before operating on specific collections. Returns array of collection objects with id, name, type, and fields.'
Document the output schema explicitly. Define what fields each collection object contains (at minimum: id, name, type, created, updated, system). This allows LLMs to extract specific collection IDs for follow-up operations.
Add optional input parameters to support filtering or pagination: consider limit (1-1000, default 50), offset (for pagination), and filter_by_name (optional substring match). This makes the tool more useful for agents working with many collections.
Implement and document the promised resources and prompts capabilities, or remove them from the capabilities declaration to avoid misleading clients.
Enhance error messages with recovery guidance. For example: 'Failed to authenticate with PocketBase: Invalid credentials. Check that PB_ADMIN_EMAIL and PB_ADMIN_PASSWORD are correct and match an admin user in the target PocketBase instance.' This helps LLMs diagnose and recover from common failures.
Add a note in the tool description about prerequisites: 'Requires a running PocketBase instance with valid admin credentials provided via environment variables (PB_URL, PB_ADMIN_EMAIL, PB_ADMIN_PASSWORD) or command-line arguments.'

Score history

Overall score trend

↑ 27 points across a rubric change (v1 → v2)

50/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-22	D	50	2026-07-28+	v2
2026-03-09	F	23	-	v1