Work-Bench
Work-Bench is New York City's B2B software seed fund, founded in 2013, backing founders building the next generation of applied AI and infrastructure software. The firm leads pre-seed and seed rounds and works with founders on go-to-market from the first customers through scaling, under the thesis that "GTM is your moat". It is NYC-based and NYC-biased, but backs founders anywhere who want to win enterprise customers in New York. Its portfolio spans three areas, AI infrastructure and developer tools, cybersecurity, and enterprise apps and vertical AI, with exits including CoreOS (Red Hat), Semmle (GitHub), Tilt (Docker), Permiso (Okta) and FireHydrant (Freshworks). Headquartered at 915 Broadway, New York. It publishes the GTM Weekly newsletter.
Work-Bench has 38 portfolio companies tracked on the APIs.io network, of which 20 (53%) publish APIs indexed here. The remaining 18 are tracked as profiles without a published API on the network.
Its highest-rated API providers on the network are FireHydrant, Socure, Arthur, Peerbound, Shipyard.
Portfolio on the Network
Portfolio companies that publish APIs indexed on APIs.io, grouped by their API Evangelist rating. Exemplar and Strong are expanded by default.
Exemplar 1 Complete, well-documented, and agent-ready
Thin 5 Limited public surface area
Properties
Cohort brief
Auto-generatedThe 20 portfolio companies with APIs in the catalog, backed by Work-Bench, scored on the Kin Score. Every figure below is computed from the catalog — nothing here is written.
| Facet | This cohort | Catalog | Difference | Scored |
|---|---|---|---|---|
| Access Clarity | 41.3 | 26.5 | +14.8 | 20 |
| Operational Transparency | 18.2 | 13.2 | +5 | 20 |
| Contract Quality | 22.4 | 17.7 | +4.7 | 20 |
| Discoverability | 60.1 | 58 | +2.1 | 20 |
| Developer Ergonomics | 25.6 | 23.8 | +1.8 | 20 |
| Contract Governance | 8.2 | 7 | +1.2 | 20 |
A facet is averaged over the members that carry it, not over the whole cohort — the “Scored” column is that count. Averaging an absent facet as zero would score our own coverage gaps as the providers’ posture.
| Artifact | This cohort | Catalog | Difference |
|---|---|---|---|
| MCP server (any) | 10% | 15% | -5 |
| MCP server (first-party) | 5% | 13% | -8 |
| Agent Skills | 0% | 0% | 0 |
| OAuth scopes | 15% | 12% | +3 |
| Security | 100% | 98% | +2 |
| Arazzo workflows | 5% | 2% | +3 |
| Governance rules | 15% | 18% | -3 |
mcp_pct counts any mcp/ artifact including ones API Evangelist derived from the provider OpenAPI; mcp_first_party_pct counts only servers the provider publishes. Prefer the latter.
- 1 Chamelio {"score" => 48.5, "band" => "agent-ready"}
- 2 ArthurAI {"score" => 42, "band" => "agent-ready"}
- 3 ZippedScript {"score" => 38.8, "band" => "agent-ready"}
- 4 Peerbound {"score" => 37.9, "band" => "agent-ready"}
- 5 VTS {"score" => 37.9, "band" => "agent-ready"}
- 6 FireHydrant {"score" => 35.1, "band" => "agent-ready"}
- 7 DevArmor {"score" => 32.9, "band" => "agent-ready"}
- 8 Petra {"score" => 26.4, "band" => "agent-aware"}
- 9 Dialpad {"score" => 23.4, "band" => "agent-aware"}
- 10 Socure {"score" => 20, "band" => "agent-aware"}
Coverage caveat. 16 of 20 scored members had under half their catalog-attributable points satisfied at all (catalog_earned / catalog_max < 0.5). They are INCLUDED in the mean above. A large share here means the cohort mean is partly a measurement of how far our enrichment reached, not only of how these providers publish.Generated from the catalog build of 2 November 2026, across 37080 providers, using the same computation served by the APIs.io cohort API. Briefs are published for rosters of 5 or more scored providers; below that a distribution is not meaningful.
Source (apis.yml)
Work with this as data
Every firm here is available over the APIs.io API and to AI agents over MCP.