Provides tools for interacting with Supabase, including querying allowlisted tables, searching and adding memory entries, creating project records, and managing workspace files.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Zynx Supabase MCP Server Search Zynx memory for MCP architecture."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Real TypeScript MCP server for connecting Zynx agents to Supabase through safe, allowlisted tools.
get_agent_health — check MCP server health
list_allowed_tables — show table allowlist and limits
query_table — read rows from allowlisted Supabase tables
search_memory — keyword search over Zynx memory table
add_memory — persist project, research, coding, or automation memory
create_project — create a project orchestration record
list_workspace_files — list files under configured workspace roots
read_workspace_file — read small text files under configured workspace roots
Related MCP server: SupaMCPBuilder
cd zynx-supabase-mcp
cp .env.example .env
npm install
npm run build
npm start.envZYNX_ENV=development
SUPABASE_URL=https://YOUR_PROJECT_ID.supabase.co
SUPABASE_KEY=YOUR_SUPABASE_SERVICE_ROLE_OR_SERVER_KEY
ZYNX_ALLOWED_TABLES=zynx_projects,zynx_project_tasks,zynx_memory,zynx_audit_logs,zynx_documents,zynx_vectors,zynx_tool_runs
ZYNX_DEFAULT_LIMIT=10
ZYNX_MAX_LIMIT=100
ZYNX_AUDIT_TABLE=zynx_audit_logs
ZYNX_ENABLE_AUDIT=true
ZYNX_WORKSPACE_ROOTS=/Users/kant/zynx-supabase-mcp
ZYNX_MAX_FILE_BYTES=200000Use .env.staging.example and .env.production.example as profile templates when those environments are created.
Option A: command line:
gemini mcp add zynx-supabase node /ABSOLUTE/PATH/TO/zynx-supabase-mcp/dist/index.jsOption B: copy config/gemini-settings.example.json into your Gemini CLI settings and replace values.
Use get_agent_health from zynx-supabase.Use list_allowed_tables from zynx-supabase.Use query_table to read 5 rows from zynx_memory.Search Zynx memory for MCP architecture.List workspace files from /Users/kant/zynx-supabase-mcp.Add this project decision to Zynx memory.Create a Zynx project named Zynx MCP Dev.Never expose SUPABASE_KEY in frontend code.
Keep .env out of git.
Keep ZYNX_ALLOWED_TABLES narrow.
Keep ZYNX_MAX_LIMIT small.
Prefer read-only database roles/RPC in production.
Enable Supabase RLS where possible.
See:
supabase/schema.example.sqlIt creates starter tables for:
zynx_audit_logs
zynx_memory
zynx_projects
zynx_project_tasks
zynx_tool_runs
zynx_documents
zynx_vectors
See docs/architecture.md for the dev-first plan that still supports staging and production.
After npm run build:
./scripts/register-gemini.shPersist project, research, coding, or automation memory in an allowlisted Supabase table.
| Name | Required | Description | Default |
|---|---|---|---|
| kind | No | note | |
| content | Yes | Memory text to persist. | |
| metadata | No | ||
| projectId | No | ||
| tableName | No | Allowlisted memory table. | zynx_memory |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description must disclose behavior, but it only says 'persist'. It fails to mention side effects (database write), failure modes (if table not allowlisted), or return values, and doesn't clarify whether this creates a new record or updates existing ones.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single concise sentence that directly states the action, with no extraneous words, making it well-structured and easy to parse.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Despite having 5 parameters and no annotations or output schema, the description only covers the basic purpose. Missing details about parameter meanings, constraints, and expected behavior make it inadequate for reliable tool invocation.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool description gives no parameter semantics; it doesn't explain 'kind', 'metadata', or 'projectId'. The input schema only describes 'content' and 'tableName', leaving 60% of parameters unexplained, and the description does not compensate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb ('Persist') and identifies a resource ('project, research, coding, or automation memory') within a 'Supabase table', clearly stating the action. However, it doesn't explicitly differentiate from sibling tools like search_memory or query_table, leaving some ambiguity about when this is the right tool.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No usage guidance is provided. The description lacks context on when to prefer this tool over search_memory or query_table, and it doesn't mention prerequisites like table allowlisting or projectId requirements.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Create a Zynx project record for orchestration, memory, research, coding, and automation work.
| Name | Required | Description | Default |
|---|---|---|---|
| name | Yes | Project name. | |
| status | No | Project status, usually active/paused/archived. | active |
| metadata | No | ||
| description | No | Optional project description. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description must carry the full burden. It only states that a record is created, with no details on permissions, side effects, idempotency, or return behavior. This is insufficient for a create operation.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Single sentence that is clear and to the point, with no redundant information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Despite moderate schema coverage, the description omits expected return values, prerequisites, and behavior on duplicates. With no output schema and no annotations, the description is too terse to fully inform an agent.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 75%, so baseline is 3. The description adds no parameter-specific meaning beyond what the schema already provides (e.g., it does not clarify expected metadata structure or status values).
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Description uses specific verb 'Create' and resource 'Zynx project record', clearly distinguishing from sibling tools focused on health, memory, queries, and file access.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Description mentions the purpose ('orchestration, memory, research, coding, and automation work') implying when a project record is needed, but lacks explicit alternatives or exclusions. No guidance on when to use other tools.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Check whether the Zynx Supabase MCP server is alive and configured.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description is the sole source of behavioral disclosure. It conveys that the tool performs a status check with no side effects, but does not describe output format, potential failure modes, or whether it requires authentication. For a simple health check, this is adequate but not rich.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, direct sentence that front-loads the action verb 'Check' and contains no redundant phrasing. It earns its place with every word conveying essential meaning.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's extreme simplicity (0 parameters, no output schema, no annotations), the description fully captures its purpose and resource. The sibling tools are all data operations, so there is no ambiguity about when this health check is relevant. The description is complete for the tool's complexity.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool has zero parameters, so the description correctly omits parameter details. Per the baseline for 0-parameter tools, a score of 4 is appropriate since the schema is empty and no additional semantic meaning is needed.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses the specific verb 'Check' and names the resource 'Zynx Supabase MCP server', with the scope 'alive and configured'. This clearly distinguishes the tool from sibling data-operation tools like query_table and search_memory, which are unrelated in purpose.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides clear context: this is a health-check tool to verify server liveness and configuration, which implies use when diagnosing server availability. It does not explicitly mention alternatives, but the context is sufficient given the stark contrast with sibling tools, so no exclusions are needed.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List Supabase tables this MCP server is allowed to read.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the burden. It states the tool lists tables the server 'is allowed to read,' implying a read-only operation and no side effects. However, it provides no details on output format, ordering, or error behavior, which is minimal but acceptable for a simple list operation.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, concise sentence that immediately states the tool's function. There is no redundancy or filler, making it highly efficient.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the low complexity (no parameters, no output schema), the description is sufficiently complete. It tells the agent exactly what the tool does and what it returns conceptually. Slight lack of usage hints prevents a 5.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool has zero parameters, so per the rubric the baseline score is 4. The description does not need to explain any parameter semantics since none exist.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb ('List') and resource ('Supabase tables this MCP server is allowed to read'), clearly distinguishing it from sibling tools like query_table which queries table content. It conveys both the action and the scope of what is returned.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage is to discover which tables are readable, but it does not explicitly state when to use this tool versus alternatives like query_table. No exclusions or prerequisites are mentioned, but the purpose is clear enough from context.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List files inside configured Zynx workspace roots for coding and automation tasks.
| Name | Required | Description | Default |
|---|---|---|---|
| depth | No | ||
| limit | No | ||
| dirPath | No | Directory to list. Defaults to the first configured workspace root. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description must disclose behavioral traits, but it only mentions the scope constraint of working within configured workspace roots. It does not discuss depth/recursion behavior, return format, or any side effects, leaving significant behavioral gaps.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, concise sentence that is front-loaded with the action verb. It contains no redundant words, though 'for coding and automation tasks' adds minimal value.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a tool with three parameters, no annotations, and no output schema, the description leaves important gaps. It does not describe what the output looks like, whether directories are included, or how depth and limit behave, making it incomplete for an agent.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is only 33% (only dirPath has a description). The tool description does not explain the 'depth' or 'limit' parameters at all, failing to compensate for the low schema coverage.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses the specific verb 'List' with the resource 'files inside configured Zynx workspace roots', clearly distinguishing it from sibling 'read_workspace_file' which reads file contents. It is concise and unambiguous.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance is provided on when to use this tool versus alternatives. It vaguely mentions 'for coding and automation tasks' but gives no explicit context, exclusions, or comparisons to siblings like read_workspace_file.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Read rows from an allowlisted Supabase table with optional equality filters.
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | Rows to return. Capped by ZYNX_MAX_LIMIT. | |
| select | No | PostgREST select string. Default: *. | * |
| filters | No | Simple equality filters: { column: value }. | |
| orderBy | No | Optional column for descending order. | |
| tableName | Yes | Allowlisted Supabase table name. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description must fully disclose behavioral traits. It states the read-only nature and the allowlist constraint, but omits important details like what happens when a table is not allowlisted, whether results are paginated, and any default limits or error behavior. This leaves an agent uncertain about operational outcomes beyond the simple read action.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single front-loaded sentence that immediately conveys the core action, resource, and constraint. It is concise, with no redundant or extraneous information, making it easy for an agent to parse quickly.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The tool has five parameters, a nested object, and no output schema or annotations. The description provides a basic purpose but is too terse to cover error handling, default behaviors, or how to manage large result sets. The schema covers parameter details, but the overall description leaves some functional gaps.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema provides descriptions for all five parameters, so the baseline is 3. The description's mention of 'optional equality filters' aligns with the filters parameter but does not add new semantic value beyond what the schema already specifies.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool reads rows from an allowlisted Supabase table with optional equality filters, using a specific verb and resource. This distinguishes it from sibling tools like list_allowed_tables, which only lists table names, and read_workspace_file, which targets files.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage for reading table data but provides no explicit guidance on when to choose this tool over alternatives such as list_allowed_tables or search_memory. The context of reading from an allowlisted Supabase table is clear, but no exclusion criteria or alternative recommendations are given.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Read a small text file inside configured Zynx workspace roots.
| Name | Required | Description | Default |
|---|---|---|---|
| filePath | Yes | File path inside one configured workspace root. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the burden. It does disclose that the operation is a read (safe) and that it is limited to small text files within configured roots. However, it lacks details about error behavior, exact size limits, or whether symbolic links are followed. The transparency is present but minimal.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, well-structured sentence. It front-loads the action and includes all essential information without redundancy. Every word contributes to the tool's purpose and constraints. Perfectly concise.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple read tool with one parameter and no output schema, the description covers the core requirements: the action, the object type, and the location restriction. It could be more complete by specifying error outcomes or the 'small' size threshold, but the essential context for invoking the tool is present. Slight gap in usage guidance is already reflected in that dimension.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema already describes the only parameter (filePath) with 100% coverage, so the baseline is 3. The description adds a note about 'small text file' and 'configured workspace roots,' but these are tool-level constraints, not parameter-specific semantics. No additional meaning is provided for filePath beyond what the schema already states.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb (Read) and a clear resource ('small text file inside configured Zynx workspace roots'). It differentiates from sibling tools like list_workspace_files (listing) and query_table (querying). The scope is clearly defined, making the tool's purpose unambiguous.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Usage context is implied rather than explicit. The description indicates what the tool does but does not state when to use it over alternatives (e.g., when to read a file vs list files or query a table). There is no mention of exclusions or prerequisites. This is consistent with a 'minimum viable' level of guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Keyword search over an allowlisted Zynx memory table using ilike.
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | ||
| query | Yes | Keyword to search in Zynx memory content. | |
| tableName | No | Allowlisted memory table. | zynx_memory |
| contentColumn | No | Column containing searchable text. | content |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure. It mentions 'allowlisted' and 'ilike' but omits details like return format, ordering, limit behavior, or what happens when no matches occur. This is minimal for a search tool.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, well-structured sentence that front-loads the primary purpose. It is concise and contains no filler, making it efficient for an agent to parse.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
There is no output schema and the description does not explain return values, ordering, or pagination. The context around 'Zynx memory' and 'allowlisted' is also unelaborated, leaving the agent with insufficient information for a tool with 4 parameters and no response schema.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 75% with descriptions for query, tableName, and contentColumn, so the schema already explains most parameters. The description adds the 'ilike' behavior but does not clarify limit semantics or expand on any parameter beyond schema text. This is adequate but not additive.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action ('Keyword search') and the resource ('allowlisted Zynx memory table'), specifying the matching mechanism ('using ilike'). This distinguishes it from sibling tools like query_table by focusing on memory content and fuzzy text matching.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides no guidance on when to use this tool versus alternatives such as query_table or list_allowed_tables. There are no explicit use cases, exclusions, or preferred scenarios, leaving the agent without contextual selection criteria.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v1.0.0add_memorycreate_projectget_agent_healthlist_allowed_tableslist_workspace_filesquery_tableread_workspace_filesearch_memoryScored across 8 tools
Each tool targets a distinct resource-action pair: health, table listing, generic query, memory search, memory write, project create, file listing, and file read. No two tools perform the same operation on the same resource.
All tool names follow a consistent verb_noun pattern in snake_case (get_, list_, query_, search_, add_, create_, read_). The style is uniform and predictable.
With 8 tools, the server covers its core capabilities without bloat. It is within the ideal 3-15 range and each tool serves a clear purpose.
The tool surface is notably incomplete: memory supports add and search but no update or delete; projects support create but no retrieval, update, or delete; workspace files support list and read but no write. These are significant gaps for a server that claims to support orchestration, memory, and automation.
MCP server for interacting with the Supabase platform
Nifty's MCP server — exposes tasks, projects, messages, and files as tools for AI agents.
Official Airtable MCP server — database and operations layer for agents.
Remote MCP server for supportsheep: run AI interviews and manage support content for your blog.
