OpenSpec

Plan mode for complex features

Winter 2026ActiveB2BArtificial IntelligenceDeveloper ToolsB2BOpen SourceSan Francisco, CA, USA

Company

OpenSpec is an open-source, spec-driven framework. We help developers plan complex features by creating lightweight spec documents that serve as durable context for their coding agents. With 20k+ GitHub stars, we’re one of the leading spec-driven frameworks, used by thousands of developers and teams to plan and ship with the coding agent of their choice. Our philosophy is simple: keep planning lightweight and iterative, and match how software actually gets built, messy, evolving, and real, without dragging teams into rigid, bureaucratic workflows.

Verdict

Medium Signal

Market Opportunity

Developer tooling for AI coding agents is a real and growing market, but ICP is still broad ('developers and teams') with no enterprise-specific targeting yet. B2B SaaS angle via Workspaces is the monetization path but hasn't launched. TAM is meaningful given the explosion of AI coding agent adoption, but OSS-to-paid conversion is historically hard.

Medium Signal

Founder Signal

Solo founder Tabish Bidiwale has ~5.5 years at Q-CTRL (Graduate → Senior SWE → Team Lead), joining early and scaling through a $113M Series B — that's legitimate hands-on engineering experience at a real company. University of Sydney CS grad (2019). Concern: solo technical founder with no apparent co-founder, and no prior exit. Relevant domain insight (saw agent misalignment at Q-CTRL) is credible but execution risk is high alone.

Medium Signal

Competition

No competitor data was returned in research, but the space includes agent-native planning tools, Cursor's built-in plan mode, and GitHub Copilot Workspace. The 'no API keys, no MCP' positioning differentiates from heavier solutions. The 28.5K stars suggest real differentiation vs. alternatives, but big players (Anthropic, GitHub, Cursor) can natively improve their planning layers and squeeze this out.

High Signal

Product

28.5K GitHub stars is exceptional organic developer traction for an open-source tool — that's a real signal, not vanity. Native support for Claude Code, Cursor, Codex, GitHub Copilot plus 16+ more tools shows real integration work shipped. However, no revenue metrics visible and the team product (Workspaces) is still 'coming soon / in development', meaning monetization is unproven.

OverallB Tier

OpenSpec has one genuinely impressive signal — 28.5K GitHub stars is top-decile organic traction for a developer tool and proves real product-market fit at the individual developer level. The core problem (AI coding agent misalignment on complex features) is real and well-articulated from direct experience at Q-CTRL. The fatal weakness right now: solo founder, no revenue, and the team/enterprise product that would actually generate money is still unbuilt. OSS-to-paid conversion is notoriously brutal, and the window to monetize before Cursor/GitHub/Anthropic natively absorb this functionality is narrow. Strong seed-stage bet if they close a co-founder and launch Workspaces fast; otherwise risks being a beloved free tool with no business.

Active Founders

Tabish Bidiwale

Founder

Building OpenSpec, the open-source spec-driven framework that makes AI coding reliable on complex work. 20k+ GitHub stars. Previously Team Lead at Q-CTRL. Joined as an early engineer and built the core platform and multiple products from seed to a $113M Series B. While at Q-CTRL, I saw coding agents recreate the same failure mode as human teams: misalignment. Existing tooling wasn’t built for that shift, which led me to building OpenSpec.

OpenSpec

TierB Tier

BatchWinter 2026

StatusActive

LocationSan Francisco, CA, USA