spec-editor Pro is in early access. Request access

Paid plugin ยท early access

spec-editor Pro

spec-editor Pro is a plugin for the open core. It adds the team that turns a requirement into reviewed, tested, released code, then watches it run and learns from the result. It works on your own subscriptions.

Features

Commercial features

The delivery team

  • Roles

    Analysts, an analysis manager, a project manager, a coder, a tester, devops and an on-duty role for incidents.

  • Who owns what

    Manager roles own the queues. A deterministic engine only moves events.

  • Independent analysis

    Analysts work on a source independently, and a second pass reconciles their results.

  • Check against the source

    The result is checked against the source: the model does not receive what the source did not say.

  • Fast path for small edits

    A typo, a label or one line takes a short route instead of the full cycle.

  • Several projects

    One installation runs several projects, each with its own engine, on shared subscriptions with quotas.

Review and release

  • Review stages decided by evidence

    Risk classes A to D, managed with spec-editor gate. A human is asked only where evidence is missing.

  • Test stand

    Every plan is deployed to a blue/green test stand, and its acceptance criteria are run there before promotion.

  • Visual testing

    Screens are accepted by a screenshot taken on the stand, not by the structure of the page.

  • Rollback

    Rollback exists for a promoted plan.

  • Remote host

    Builds and stands can run on a remote host. Coders stay on the machine where the subscription is logged in.

Monitoring

  • Log signatures

    Repeating errors in the logs are recognised by their signature and become incidents.

  • On-duty role

    An on-duty role sorts incidents on an inexpensive model and escalates what looks serious.

  • Back to the requirement

    An incident is bound to the requirement whose code wrote the record, and goes through the same analysis as any other source.

Cost control

  • Difficulty classifier

    Classifies how hard a task is.

  • Ladder of models

    From cheap to strong.

  • Cost attribution

    Per source and per role.

  • Budget guards

    Keep spend within a budget you set.

  • Pool of subscription accounts

    Work fails over to another account when one hits its limit.

  • Automatic policies

    Routine operator decisions are taken by policy, with a dry run first.

Self-improvement

  • KPIs evaluated

    The KPIs of your model are computed from the events of the pipeline, with their history.

  • Observer

    Once a day an observer looks for KPI deviations and proposes hypotheses.

  • Experiments

    Parameters of the pipeline change through declared experiments with a decision rule. Changes to code go through you.

  • Roles that learn

    Roles keep lessons from finished runs in role memory.

  • Prompts tuned by results

    The instructions of a role are refined from the outcomes of its runs, and a learned rule is checked by replay before it is kept.

Feedback from the running product

  • Breakpoints

    Visual comments right in your running product. A small JavaScript snippet is added to it; a click on an element of the screen creates a source bound to the requirement of that element.

Access control

  • Permissions per role

    Each role of the delivery team gets the tools and permissions its skills call for, and no more.

  • A fenced coder

    The coder works in the shell under a deny list of dangerous commands.

  • Named human decisions

    A review stage or a source is approved by a person, and the approval is recorded with who gave it.

Existing code

  • Reengineering

    Requirements restored from your existing codebase. Started from the interface.

  • Coverage

    spec-editor coverage shows coverage of requirements by code, both ways.

  • Local semantic search over code

    Works without an external service.

  • Syntax parsers

    Python, TypeScript, JavaScript, Go, Java, Kotlin and Rust are parsed, so agents see functions and classes, not lines of text.

Documents and demos

  • User manual

    Assembled from the scenarios and screens of the model.

  • Demo video

    A chosen scenario is run on a chosen stand and recorded, with chapters and a poster.

In the interface

  • Report a bug

    From the interface.

  • Chat with an agent

    In the context of an element.

  • Opt-in statistics

    A protocol for sharing anonymous aggregate statistics exists. It is opt-in.

Use

Scenarios that become available

Turn a source into checked code

What happens

  1. Analysts work on the source independently, a second pass reconciles them, and the result is checked against the source.
  2. The requirement gets acceptance criteria, then a plan, then code written by a coder and checked by a tester.
  3. The plan runs on a test stand and passes a review stage before promotion.

Uses: delivery team, check against the source, test stand, review stages

Review a change by evidence

What happens

  1. A change gets a risk class from A to D.
  2. The review stage decides from the evidence it has.
  3. A human is asked only where evidence is missing.

Uses: review stages, spec-editor gate

Try every plan before it goes live

What happens

  1. The plan is deployed to a blue/green test stand.
  2. Its acceptance criteria are run there.
  3. Only then is it promoted; rollback exists.

Uses: test stand, acceptance run, rollback

Turn a click on the product into a requirement change

What happens

  1. A snippet in the running product lets you click an element of the screen.
  2. The click creates a source bound to the requirement of that element.
  3. The source goes through the same analysis as any other.

Uses: breakpoints, delivery team

Give an existing codebase a model

What happens

  1. Start reengineer from the interface.
  2. A model is restored from the existing code.
  3. Check what is covered with spec-editor coverage, both ways.

Uses: reengineer, coverage, local semantic search

See what a requirement cost and keep spend in budget

What happens

  1. Cost is attributed per source and per role.
  2. A classifier rates the difficulty and a ladder of models picks from cheap to strong.
  3. Budget guards keep spend within the budget you set.

Uses: cost attribution, classifier, ladder of models, budget guards

Keep working when a subscription hits its limit

What happens

  1. Several subscription accounts form a pool.
  2. When one hits its limit, work continues on another account.

Uses: pool of subscription accounts with failover

Watch the running product and fix what breaks

What happens

  1. Repeating errors in the logs become incidents.
  2. An on-duty role sorts them and escalates what looks serious.
  3. The incident arrives as a source bound to the requirement, and the fix goes through the same cycle.

Uses: log signatures, on-duty role, hierarchical logging, delivery team

Let the pipeline improve itself

What happens

  1. You set KPIs in the model.
  2. An observer finds deviations and proposes hypotheses.
  3. Experiments change parameters and keep what works; roles learn from finished runs.

Uses: KPI evaluation, observer, experiments, roles that learn

Hand a client a manual and a demo

What happens

  1. The user manual is assembled from the scenarios and screens of the model.
  2. A scenario is run on a stand and recorded as a demo video.

Uses: user manual, demo video, test stand

Build and test away from your machine

What happens

  1. Builds and test stands run on a remote host.
  2. Coders stay on the machine where the subscription is logged in.

Uses: remote host, test stand

Editions

Open core and Pro side by side

CapabilityOpen corespec-editor Pro
Model and tools
Requirements as Markdown files in git, with acceptance criteria and validationIncludedIncluded
Sources from Jira, GitHub Issues, Slack, Discord and Telegram; import of OpenSLO filesIncludedIncluded
Export: SRS in the style of IEEE 830, styled HTML, TRLCIncludedIncluded
Web interface, run log, run-time trace map and event graphIncludedIncluded
Editor extensions for VS Code, JetBrains IDEs and ZedIncludedIncluded
Hierarchical logging by the tree of requirementsIncludedIncluded
KPIs as elements of the modelIncludedIncluded
Model providers: Claude Code CLI, LiteLLM, OllamaIncludedIncluded
Command line26 commands33 commands
MCP serverAbout 70 toolsAbout 115 tools
Delivery
Analysts, analysis manager, project manager, coder, tester, devops, on-duty roleNot includedIncluded
Independent analysis, a reconciling second pass and a check against the sourceNot includedIncluded
Review stages with risk classes A to DNot includedIncluded
Test stand (blue/green) with acceptance run before promotion; rollbackNot includedIncluded
Visual testing: screens accepted by screenshotNot includedIncluded
Monitoring: incidents from log signaturesNot includedIncluded
Self-improvement: KPI evaluation, observer, experiments, roles that learnNot includedIncluded
Remote host for builds and standsNot includedIncluded
Cost control: classifier, model ladder, attribution, budget guards, account pool with failover, automatic policiesNot includedIncluded
Fast path for small editsNot includedIncluded
Several projects in one installation, with shared subscriptionsNot includedIncluded
User manual and demo video from scenariosNot includedIncluded
BreakpointsNot includedIncluded
Reengineer and spec-editor coverageNot includedIncluded
Local semantic search over codeNot includedIncluded
Report a bug and chat with an agent from the interfaceNot includedIncluded

Ownership

What stays yours

  • Your subscriptions or tokens do the work. We do not sell tokens.
  • Your code is not sent to us.
  • The model files are plain Markdown in your repository. If the licence ends, the open core keeps working with them.

Roadmap

Coming soon

  • Self-service purchase and licence keys, with the Pro package as a download
  • Account area on the site
  • Hosted view of a model

A separate product, later

The federation connector links the projects of one organisation. It will be sold apart from spec-editor Pro.

Early access

Ask for early access

Purchase is not open yet. Leave an e-mail address and we will write when it opens. Prices are early and may change. Installation and setup of the plugin are described in its documentation.