Agent projectby CannerHEU
Better agent tooling for Databricks
Integration gap · Measured demandOverall 55.9 / 100 DERConfidence Medium DERCalculated
573 public GitHub issues mention "Databricks mcp" (checked 2026-11-21) SRC, while 12 indexed products mention Databricks DER. On the six weighted components it scores 55.9 of 100, confidence Medium, as of 2026-11-21.
Score breakdown
Contribution = weight × component score. The contributions add up to the overall score.
| Component | Weight | Score | Score (0–100) | Contribution | Kind |
|---|---|---|---|---|---|
| Demand | 0.25 | 46.0 | 0.25 × 46.0 = 11.50 | Measured | |
| Competition gap | 0.20 | 46.0 | 0.20 × 46.0 = 9.20 | Measured | |
| Monetisation | 0.20 | 85.0 | 0.20 × 85.0 = 17.00 | Heuristic | |
| Growth (relative) | 0.15 | 15.0 | 0.15 × 15.0 = 2.25 | Measured | |
| Feasibility | 0.10 | 80.0 | 0.10 × 80.0 = 8.00 | Heuristic | |
| Distribution | 0.10 | 79.0 | 0.10 × 79.0 = 7.90 | Measured | |
| Overall | 1.00 | 55.85 → 55.9 |
Solid bars are measured components; dashed outlines are rules of thumb. The full formula is on the radar page.
Confidence: Medium
Rule: Medium when at least 50 GitHub issues mention the integration and at least 10 indexed products mention it; otherwise Low. Nothing is rated High.
What would raise it
- Medium is the highest level the rules assign. High would need a second, independent demand source (search volume or marketplace installs), which is not connected.
The opportunity
All text in this section is generated by rule from the record. It is a starting point, not research.
- Description
- Developers are asking for Databricks access from AI agents more than the current supply of well-maintained tools would suggest.DER
- Problem
- Agents need to read from and act on Databricks. Only some of the existing tools are both documented and actively maintained.DER
- Target users
- Teams that already pay for Databricks and want agents to work inside itHEU
- Product form
- MCP server (optionally bundled as a plugin with skills)HEU
- Revenue model
- Hosted/remote server with per-seat or usage pricingHEU
- Complexity
- Low to mediumHEU
- Competition level
- MediumDER
- Subject
- Databricks (integration)
Evidence for
4 row(s) from opportunity_evidence.
Competition gapMeasuredDERvalue 10
12 indexed products mention Databricks; 10 of them have a complete record and a repository pushed in the last 180 days.
MonetisationHeuristicHEU
Databricks is sold mainly to businesses on paid plans, which suggests buyers with budget; "data-warehouse" tooling is a category companies commonly pay for. This is a rule of thumb, not revenue data.
FeasibilityHeuristicHEU
Databricks has a public API (documentation linked).
DistributionMeasuredDERvalue 37
37 of 47 tracked agent platforms document MCP support, so one MCP server can reach all of them.
Evidence against
4 row(s) from opportunity_evidence.
DemandMeasuredDERvalue 573
573 public GitHub issues contain the phrase "Databricks mcp" (checked 2026-11-21); that is the 46th percentile of the 108 integrations tracked. Issue mentions measure developer discussion, not buyers.
Growth (relative)MeasuredDERvalue 3
3 products launched in the last 90 days versus 2 in the previous 90; the whole catalogue went from 14427 to 45454, and the score is relative to that.
Competition gapMeasuredDER
One project holds 97% of the 18,885 GitHub stars in this segment, so there is an established leader to displace.
Competition gapMeasuredDERvalue 2
2 indexed product(s) are published under a namespace matching Databricks itself (Databricks Field Guide by TechFabric, databricks). The gap score was reduced by 25 points for that.
Existing products
All Databricks productsThe most notable profiled products mentioning Databricks (up to 8). 12 indexed in total.
Pluginby databricksSRC
Pluginby oracle-samplesSRC
Pluginby airtableSRC
Pluginby altimateaiSRC
MCP serverby bricksentrySRC
MCP serverby mcpsmithsSRC
MCP serverby camber-mcp.cambercloud.comSRC
Where it could be distributed
All platforms37 of 47 tracked agent platforms document MCP support CUR, so one MCP server could reach all of them. No pairing has been tested.
ContinueOpenCodeClaude CodeCodexGemini CLIZedOpenHandsClineWarpgooseQwen CodeRoo Code+25 more
Channels: Official MCP Registry listing, plugin marketplaces for Claude/Cursor/Codex, and the integration vendor's own app directory where one exists. HEU
Data coverage
- Indexed products
- 12DER
- With a full profile
- 8+DER
- Launches, last 90 days
- 3 vs 2 in the 90 beforeSRC
- Demand measured
- SRC
- Segment calculated
- Score calculated
Risks and counterarguments
- The vendor may ship (or already ships) an official Databricks server, which usually becomes the default.
- Issue mentions measure developer discussion, not willingness to pay.
- 573 public GitHub issues contain the phrase "Databricks mcp" (checked 2026-11-21); that is the 46th percentile of the 108 integrations tracked. Issue mentions measure developer discussion, not buyers.
- 3 products launched in the last 90 days versus 2 in the previous 90; the whole catalogue went from 14427 to 45454, and the score is relative to that.
- One project holds 97% of the 18,885 GitHub stars in this segment, so there is an established leader to displace.
- 2 indexed product(s) are published under a namespace matching Databricks itself (Databricks Field Guide by TechFabric, databricks). The gap score was reduced by 25 points for that.
- The catalogue already holds 2 products for Databricks published by an account named after it: databricks (databricks), Databricks Field Guide by TechFabric (databricks.techfabric.com). That may be the vendor's own server; an account name is not proof.
Assumptions
What has to be true for the score to mean anything.
- Mentions of "Databricks mcp" in GitHub issues reflect developers wanting Databricks access from agents.
- Products that mention the integration in their description or package metadata are the competing supply; ones that do not name it are missed.
- "Well-maintained" means data completeness ≥ 0.6 and a repository push in the last 180 days.
- Percentiles compare this integration only with the other integrations tracked, not with the wider market.
- A platform that documents MCP support can run any MCP server; no pairing has been tested.
How to verify this yourself
- Run the demand query on GitHub: https://github.com/search?q=%22Databricks%20mcp%22%20in%3Atitle%2Cbody%20is%3Aissue&type=issues. Compare the count with 573 on 2026-11-21; read a sample of issues to see whether they ask for agent access or merely mention it.
- Check the supply on the Databricks page and search the official MCP Registry and the vendor's own app directory for an official server.
- Download the record with all evidence: /api/dataset/opportunities.json.
Page data last updated .