Manage Coalesce Transform workspaces, nodes, pipelines, and runs programmatically via MCP. Built for data engineers and DevOps teams integrating Coalesce into AI-assisted workflows.
io.github.jessemarshall/coalesce-transform
Local install
STDIO
No auth required
How models use it and what it is built for.
Manage Coalesce Transform workspaces, nodes, pipelines, and runs programmatically via MCP. Built for data engineers and DevOps teams integrating Coalesce into AI-assisted workflows.
Local install — runs as a subprocess.
Configuration this server reads at startup.
Bearer token from the Coalesce Deploy tab. Optional when `~/.coa/config` provides a `token`.
Selects which `~/.coa/config` profile to load. Defaults to default.
Region-specific base URL. Defaults to https://app.coalescesoftware.io (US).
Fallback org ID for cancel-run. Also readable from `orgID` in the active ~/.coa/config profile.
Local repo root for repo-backed tools and pipeline planning. Also readable from `repoPath` in the active ~/.coa/config profile.
Base directory for the local data cache. When set, cache files are written here instead of the working directory. Also readable from `cacheDir` in the active ~/.coa/config profile.
JSON size threshold before auto-caching to disk. Defaults to 32768.
In-memory lineage cache TTL in milliseconds. Defaults to 1800000.
Max outbound API request body size. Defaults to 524288.
When `true`, hides all write/mutation tools during registration. Only read, list, search, cache, analyze, review, diagnose, and plan tools are exposed. Defaults to false.
Directory for customizable AI skill resources. When set, reads context resources from this directory and seeds defaults on first run. Users can augment or override any skill.
Snowflake account identifier (e.g., `abc12345.us-east-1`). Required by the local `coa` CLI and `coa doctor`; not used by the MCP's REST run path. (required for run tools)
Snowflake account username (required for run tools)
Path to PEM-encoded private key (required if SNOWFLAKE_PAT not set)
Snowflake Programmatic Access Token (alternative to key pair)
Passphrase for encrypted keys
Snowflake compute warehouse (required for run tools)
Snowflake user role (required for run tools)
Where to find authoritative docs and source for Jessemarshall Coalesce Transform.
Paste any of these into Agent Studio after connecting Jessemarshall Coalesce Transform.
Common questions about connecting and running Jessemarshall Coalesce Transform.
How do I authenticate this MCP server with Coalesce?
Set COALESCE_ACCESS_TOKEN to a Bearer token from your Coalesce Deploy tab, or store it in ~/.coa/config under the active profile. The server reads from config by default if no env var is set.
Can I use this server across multiple Coalesce regions?
Yes. Set COALESCE_BASE_URL to your region-specific endpoint (defaults to US at https://app.coalescesoftware.io). You can also switch profiles via COALESCE_PROFILE to load different region configs.
What happens when I query large lineage graphs?
The server caches lineage in memory for 30 minutes (configurable via COALESCE_MCP_LINEAGE_TTL_MS) and auto-caches large JSON responses to disk when they exceed 32 KB (configurable via COALESCE_MCP_AUTO_CACHE_MAX_BYTES).
How do I lock this server into read-only mode?
Set COALESCE_MCP_READ_ONLY=true to hide all write and mutation tools during registration. Only read, list, search, cache, and analysis tools remain available.
What should I do if API requests are hitting size limits?
Check COALESCE_MCP_MAX_REQUEST_BODY_BYTES (defaults to 512 KB). If your payloads exceed this, increase the limit or break requests into smaller chunks. Set COALESCE_REPO_PATH if using repo-backed tools for better planning.
MCP Playground runs 10,000+ hosted MCP servers — GitHub, Linear, Notion, Stripe, Sentry and more — across Claude, GPT, Gemini, DeepSeek and 30+ AI models. Compare model answers side-by-side, save agent presets, share runs. Zero install.
Open Agent Studio