OpenSpec
OpenSpec is an open-source, spec-first framework that keeps AI coding agents building the right thing — with verify and archive built in.
Free, local, MIT-licensed, and already in front of 265,000 developers a month — the adoption is real, not marketing. The /opsx:verify and /opsx:archive pair is the part worth trusting: most agent planning tools stop at the plan and never confirm the result. Pick it if you run agents on a codebase you intend to keep; skip it if you want a hosted dashboard today, since the Cloud Agent is brand new.
Verified 4d ago · liveness 77/100 · cite: rightaichoice.com/tools/openspec
- Teams running multiple AI coding agents that need one shared, portable plan
- Engineers on large brownfield codebases where context dies between chat sessions
- Reviewers who want evidence the implementation matched the stated requirements
- Organizations that need specs version-controlled in Git rather than trapped in a vendor tool
- Solo devs writing throwaway scripts where the propose/apply cycle is pure overhead
- Teams that need a hosted GUI or dashboards today
- Anyone requiring real-time collaborative spec editing today
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip OpenSpec if you want a hosted GUI with browser-based dashboards and real-time collaborative spec editing, or if you're writing throwaway scripts where the propose/apply/verify cycle is pure overhead.
OpenSpec itself costs nothing, but every /opsx command consumes tokens from your AI coding assistant's own plan — a long propose/apply/verify loop bills against Claude Code, Cursor, or Copilot, not against OpenSpec.
OpenSpec is free and MIT-licensed, with no paid tier for the core CLI — the real spend is the AI coding assistant you already pay for, whether that's Claude Code, Cursor, or GitHub Copilot. Compared with commercial spec and planning layers that bundle a hosted dashboard into a per-seat subscription, OpenSpec's cost is zero and its ceiling is your agent bill and the time to maintain the spec directory.
In short
OpenSpec — OpenSpec is an open-source, spec-first framework that keeps AI coding agents building the right thing — with verify and archive built in. Best for Teams running multiple AI coding agents that need one shared, portable plan, Engineers on large brownfield codebases where context dies between chat sessions, Reviewers who want evidence the implementation matched the stated requirements. Free to use.
What's new in OpenSpec
Checked 4 days agoAcross the latest 5 updates: 4 feature updates and 1 launch.
OpenSpec Cloud Agent introduced
OpenSpec launched a Cloud Agent, moving spec-driven workflows from the local CLI to a hosted agent.
v1.11.0: Spec Diffs & Batch Status
v1.11.0 adds spec diff visualization and batch status reporting for tracking multiple spec changes at once.
v1.10.0: Zed support, quieter installs
v1.10.0 adds Zed editor support and reduces installer output noise.
v1.9.0: Command Code & safer specs
v1.9.0 integrates Command Code and tightens spec validation safety.
v1.8.0: More agents, sturdier archives
v1.8.0 widens agent support and hardens the archive workflow.
What people actually say about OpenSpec — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
73 mentions across 4 sources (Hacker News, YouTube, GitHub, Lemmy) · researched Aug 27, 2026.
Average across the 4 sources that answered — each source counts once, not each post.
- +Persistent specs survive chat sessions and context loss — a huge win.
- +Agent-agnostic: works with Claude Code, Cursor, Codex, Copilot, and more.
- +Free and open-source, with 66k+ stars and no API keys needed.
- +Brownfield-first: drops into existing codebases without upfront spec generation.
- +Spec deltas clearly show requirement changes between proposals.
- −CLI-only; no hosted or GUI offering yet, limiting accessibility.
- −Generates a lot of markdown docs that become a maintenance burden.
- −Overkill for small projects or quick one-off changes.
- −Couples specification with execution, which can weaken specs.
- −Team features like workspaces and multi-repo planning are incomplete.
- • No direct costs, but users spend time learning the workflow and maintaining generated docs
- • Potential costs from token usage in AI assistants during the iterative spec refinement process
Viability Score
How well maintained and how widely used is OpenSpec? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- Spec-driven development with living spec deltas checked into Git
- /opsx:explore maps the problem and reads the codebase before you commit
- /opsx:propose drafts proposal.md, specs/, design.md, and tasks.md
- /opsx:apply implements tasks directly from the approved specification
- /opsx:verify checks the implementation matches the written spec
- /opsx:archive files completed changes for a durable record
- Spec diff visualization showing exactly how requirements changed (v1.11.0)
- Batch status tracking requirement changes across multiple specs (v1.11.0)
- Native slash-command integration across a long directory of AI coding agents
- AGENTS.md fallback so agents without slash support still read the spec
- Brownfield-first: adopt on existing codebases with no upfront spec backlog
- Specs organized by capability as version-controlled Markdown
- CLI install via npm, pnpm, yarn, bun, Homebrew, Deno, or Nix
- Local-first CLI with no hosted account or API keys required
- Cloud Agent for running the spec-driven workflow as a hosted agent (announced 2026-08-28)
About OpenSpec
OpenSpec is a lightweight, MIT-licensed spec framework from Fission-AI for capturing what you want built in a reviewable document your coding agent can actually follow. Instead of a prompt that evaporates when the chat closes, you write a spec inside your Git repo and OpenSpec keeps it in sync as the work evolves. The vendor reports 68.0k GitHub stars, a new spec created every two seconds, and more than 265,000 developers using it monthly. The workflow runs through five slash commands: /opsx:explore maps the problem and reads the codebase with you; /opsx:propose drafts proposal.md, specs/, design.md, and tasks.md; /opsx:apply implements the tasks from the spec; /opsx:verify checks that the implementation matches what was written down; /opsx:archive files the completed change. Specs sit as version-controlled Markdown organized by capability, and AGENTS.md fallback lets agents without slash-command support still read them. Recent releases push the review side further. v1.11.0 added spec diff visualization and batch status tracking so requirement changes across multiple specs are reviewable next to code; v1.10.0 brought Zed support and quieter installs; v1.9.0 added Command Code and tightened spec validation; v1.8.0 widened agent support and hardened the archive workflow. The current line is v1.13.0. On 2026-08-28 Fission-AI introduced a Cloud Agent that moves the workflow from the local CLI to a hosted agent. It is brownfield-first, so existing codebases work without an upfront documentation push, and installation is a single command through npm, pnpm, yarn, bun, Homebrew, Deno, or Nix. Compatibility spans a long directory of agents — Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Zed, Amazon Q Developer and dozens more — so the same spec moves with you when you switch tools. Against plan-mode transcripts or vendor-locked planning features, OpenSpec's argument is that the spec lives in your repo, survives the session, and gets checked rather than assumed.
Behind the Verdict
The pitch that lands is the durability. A plan-mode transcript dies with the chat window; a spec in your repo does not. That matters most on brownfield work, where the cost of an agent misreading three-year-old conventions is measured in review cycles, not minutes. We'd reach for OpenSpec when more than one agent touches the same codebase and the plan needs to outlive whichever tool generated it. The five-command loop is the product. /opsx:explore before /opsx:propose is the part teams skip and then regret — reading the codebase first is what stops a spec from describing a system that does not exist. Where it bites is ceremony: for a one-file script, propose/apply/verify/archive is four steps of overhead for twenty lines of code. Be honest about that threshold before you impose it on a team. The v1.11.0 spec diffs and batch status are the features I'd point a skeptical reviewer at. Requirement changes rendered as a diff, next to the code diff, is what a real review needs — and batch status is the answer to the obvious objection that specs drift the moment you have twelve of them in flight. v1.8.0 through v1.10.0 are mostly plumbing: more agents, sturdier archives, quieter installs, Zed support. Useful, not headline material. Agent coverage is the other quiet advantage. The compatibility list runs from Claude Code and Codex to Cursor, Copilot, Gemini CLI, Zed, Amazon Q Developer and dozens more, with an AGENTS.md fallback for anything that lacks slash commands. The same spec survives an agent switch, which is not a thing vendor-native planning features can claim. Cost is not the deciding factor here. The CLI is free and open source, no hosted account and no API keys, and the sources do not show a published price for the newer Cloud Agent. Treat the Cloud Agent as
Researching OpenSpec? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas OpenSpec actually fits — and what changes day-one when you adopt it.
Run /opsx:explore on an unfamiliar service to map how authentication actually works, then /opsx:propose to draft proposal.md, specs/, design.md, and tasks.md before touching code.
Outcome: The plan becomes a reviewable Markdown diff in the same pull request as the implementation, instead of a chat transcript nobody can find later.
Commit the spec to Git, let one developer work it in Claude Code and another in Codex, then run /opsx:verify against the approved spec before merging.
Outcome: Both agents work from one portable spec, and the reviewer sees explicit evidence the implementation matched the stated requirements.
Use v1.11.0 spec diff visualization and batch status to see exactly how requirements changed across several in-flight specs, then /opsx:archive the completed changes.
Outcome: Requirement churn is visible next to the code that caused it, and finished changes leave a durable archived record.
Use Cases
- Define living specs for AI-assisted feature development on existing codebases
- Generate change proposals with task breakdowns before writing code
- Review requirement deltas alongside code changes to catch misalignment
- Onboard new developers by browsing specs organized by capability
- Integrate structured specs into any Git repository for persistent context
- Switch between coding agents without losing requirement context
- Explore fuzzy ideas via /opsx:explore before committing to a proposal
- Track several in-flight spec changes at once with batch status (v1.11.0)
Limitations
- OpenSpec is a free, MIT-licensed CLI framework that requires Node.js 20.19.0 or newer and runs inside a supported AI coding assistant (Claude Code, Cursor, Codex, GitHub Copilot, Gemini CLI, OpenCode, Amazon Q Developer, Zed, and others).
- It is installed globally via npm, pnpm, yarn, bun, Homebrew, Deno, or Nix, and the vendor notes the install method that runs inside an AI assistant is new and can have varying results depending on the model used.
- The workflow depends on an external AI coding tool from the compatibility directory, so you need one already, and some IDE integrations cannot run the shell-based install prompt.
- A hosted OpenSpec Cloud Agent was announced on 2026-08-28.
as of 2026-10-04
Verification history
We have re-verified OpenSpec 10 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Showing the 6 most recent of 10 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published OpenSpec tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Open Source
$0/mo
Ideal for
Individual developers and teams who already run a supported AI coding assistant and want spec-driven planning without a hosted account, API key, or per-seat fee.
What this tier adds
Starting tier: MIT-licensed and free, covering all five /opsx commands, specs in your own Git repository, and the supported agent directory.
Where the pricing makes sense
The company stage and team size where OpenSpec's pricing actually pencils out — and where peers do it cheaper.
OpenSpec is free and MIT-licensed, with no paid tier for the core CLI — the real spend is the AI coding assistant you already pay for, whether that's Claude Code, Cursor, or GitHub Copilot. Compared with commercial spec and planning layers that bundle a hosted dashboard into a per-seat subscription, OpenSpec's cost is zero and its ceiling is your agent bill and the time to maintain the spec directory.
Setup time & first value
How long it actually takes to get something useful out of OpenSpec — broken out by persona, not the marketing-page minute.
If Node.js 20.19.0 or newer is already installed, the global CLI install is a single npm, pnpm, yarn, bun, Homebrew, or Nix command and takes under a minute. Then run openspec init in the project. Budget 15-30 minutes total to run init, confirm your assistant is on the supported list, and walk through /opsx:explore on a real feature. Prompting your assistant to install it is faster but the vendor
Switching to or from OpenSpec
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From ad-hoc chat prompts: run openspec init in the repo, then use /opsx:explore and /opsx:propose to turn the next feature request into proposal.md, specs/, design.md, and tasks.md.
- →From a docs/wiki spec folder: move capability-organized Markdown into the specs/ directory so the same spec is version-controlled in Git alongside the code.
- →From another agent's plan mode (Cursor, GitHub Copilot): keep the plan artifact but commit it as an OpenSpec change so /opsx:verify can check the implementation against it.
- →From no process at all: adopt on the existing brownfield codebase without a documentation sprint — OpenSpec is brownfield-first.
- ↗To a built-in agent plan mode: the specs are plain Markdown in your repo, so you can stop running /opsx commands and keep reading the files in Cursor or Copilot directly.
- ↗To another spec framework: because specs are version-controlled Markdown organized by capability, they travel with the repository rather than being locked in a vendor tool.
- ↗To nothing: uninstall with openspec completion uninstall followed by removing the global package — the documented steps do not touch your source code.
Integrations
Resources & Guides
- Documentationopenspec.dev
Docs · OpenSpec
Full product docs from openspec.dev
- Resourceopenspec.dev
Changelog · OpenSpec
Helpful link from openspec.dev
- Resourceopenspec.dev
Blog · OpenSpec
Helpful link from openspec.dev
- Resourceraw.githubusercontent.com
Install · OpenSpec
Helpful link from raw.githubusercontent.com
- Resourcegithub.com
OpenSpec · OpenSpec
Helpful link from github.com
Tutorials & Learning
YouTube returned 6 videos for “OpenSpec”, and we withheld 6: 6 could not be judged, because “OpenSpec” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about OpenSpec.
Tools that pair well with OpenSpec
Common stack mates teams adopt alongside OpenSpec, with the specific reason each pairing earns its keep.
OpenHands
OpenHands runs autonomous coding agents triggered by GitHub, Slack, Jira, or schedules to review PRs and fix CI.
Continue
Open-source AI coding agent for VS Code and JetBrains — free to fork since Cursor acquired Continue in January 2026.
Kiro
Kiro turns prompts into specs, validates them, then ships with parallel AI coding agents
Featured Head-to-Head Comparisons
Openspec vs Spider Cloud
Choose OpenSpec if your priority is aligning AI coding assistants with human intent through structured specs and you need free, open-source tooling. Choose Spider Cloud if you need fast, reliable web data extraction for AI agents or RAG, with pay-as-you-go pricing and recent AI Studio and browser automation capabilities.
Openspec vs Voyage Ai
Choose Voyage AI if your bottleneck is retrieval accuracy on domain-rich data (finance, legal) and you have budget for enterprise pricing. Choose OpenSpec if you want to make AI coding assistants more reliable via version-controlled specs—free and open-source. They solve different problems: retrieval versus code generation alignment.
Openspec vs Temporal Ai
If your pain point is AI agents going rogue and missing spec, OpenSpec is a lightweight, free spec layer that tames context drift. If your pain point is unreliable multi-step AI workflows that crash and lose progress, Temporal is a battle-tested durable execution AWS-level platform. For most AI teams combining agents with complex dependencies, both tools are complementary: OpenSpec for requirements, Temporal for execution. But if you had to pick one based on today's needs, Temporal's proven production deployments and recent billing improvements give it an edge for serious reliability, while OpenSpec is a no-brainer for any team that writes specs at all.
Alternatives to OpenSpec
View all
Frequently Asked Questions
Best-of guides
Used OpenSpec? Help shape our editorial sentiment research.