Open Airchor
AIRCHOR PLATFORM VISION

The control layer for agentic software development.

Airchor is building the orchestration layer between project intelligence and software execution—coordinating context, decisions, approvals, capabilities, agents, verification, evidence, and delivery.

The live MVP provides the smaller planning and project-intelligence foundation beneath that direction: persistent projects, structured context, readiness-labelled documents, roadmaps, assessments, and handoff drafts.

Open Airchorapp.airchor.comExplore MVP v1.0Sign inapp.airchor.com

MVP v1.0 is live with invitation-based access.

PLATFORM VISION

The Airchor orchestration field

  1. Project context
  2. Airchor control layer
  3. AgentTasks
  4. Execution
  5. Verification
  6. Delivery
Project intelligence passes through a governed control layer before bounded execution, independent verification, and reviewable delivery.
Read visual as text
  1. Project context
  2. Airchor control layer
  3. AgentTasks
  4. Execution
  5. Verification
  6. Delivery

Context → control layer → AgentTasks → execution → verification → delivery

THE CONTROL PROBLEM

Software can move faster than its project state

Agentic development is not only a code-generation problem. It is a coordination, authority, evidence, and verification problem across the full software lifecycle.

Context fragments

Requirements, prompts, repository facts, decisions, risk, and acceptance criteria drift across tools and sessions.

Authority stays implicit

An agent may be able to change files or call tools without a durable record of what was approved, excluded, or bounded.

Completion is self-reported

A tool can say the task is complete while scope, tests, builds, diffs, artefacts, and delivery state remain independently unchecked.

VISION

The agentic development journey

  1. Understand
  2. Structure
  3. Decide
  4. Approve
  5. Execute
  6. Verify
  7. Deliver
Airchor’s planned journey preserves project intent as work moves from understanding through governed delivery.
Read visual as text
  1. Understand
  2. Structure
  3. Decide
  4. Approve
  5. Execute
  6. Verify
  7. Deliver

Understand → structure → decide → approve → execute → verify → deliver

AIRCHOR CONTROL LAYER

Coordinate the systems around software work

Airchor is designed to sit above specialist tools—not replace them—and preserve intent, scope, permission, evidence, and the recommended next action.

  1. Preserve the project context and exact decision state.
  2. Turn approved findings into bounded AgentTasks and immutable plans.
  3. Grant only the capabilities required for the approved task.
  4. Coordinate specialist agents, repositories, runners, tests, and delivery tools.
  5. Verify outcomes independently and return evidence to the durable project record.

CONTEXT TO CONTROL

Structure the intelligence before granting execution

Project context becomes useful when sources, assumptions, risk, decisions, and confidence remain visible—and can be bound to a reviewed objective.

PLANNED

From fragmented context to project intelligence

  • Prompts
  • Documents
  • Evidence
  • Assumptions
  • Risk
  • Decisions
STRUCTUREProject intelligence
Prompts, documents, evidence, assumptions, risk, and decisions become a structured project record that can support governed work.
Read visual as text
  1. Prompts
  2. Documents
  3. Evidence
  4. Assumptions
  5. Risk
  6. Decisions
  7. Project intelligence

Every conclusion retains its source, confidence, boundary, and next decision.

GOVERNED AGENTTASKS

Approval belongs inside the execution lifecycle

A future Airchor AgentTask begins from reviewed project state, not an empty prompt. Scope, exclusions, acceptance criteria, capabilities, and stop conditions are explicit.

PLANNED WORKFLOW

Governed AgentTask lifecycle

  1. Finding
  2. AgentPlan
  3. Human approval
  4. Capability grant
  5. Execution
  6. Independent verification
A finding becomes a plan, passes an explicit human approval and capability gate, then moves through execution and independent verification.
Read visual as text
  1. Finding
  2. AgentPlan
  3. Human approval
  4. Capability grant
  5. Execution
  6. Independent verification

Finding → plan → approval → capability → execution → verification

PLANNED GOVERNANCE

Capabilities stay behind explicit boundaries

Repository, file, terminal, network, secret, database, pull-request, and deployment actions require task-scoped authority. The planned model is default deny.

These controls describe the platform direction. They are not current MVP v1.0 functionality.

PLANNED GOVERNANCE

Authority stays explicit and task-scoped

APPROVEDAgentTaskTask-scoped grants
  • RepositoryExplicit permission
  • FilesExplicit permission
  • TerminalExplicit permission
  • NetworkExplicit permission
  • SecretsExplicit permission
  • DatabaseExplicit permission
  • Pull requestsExplicit permission
  • DeploymentExplicit permission
Repository, file, terminal, network, secret, database, pull-request, and deployment capabilities sit behind explicit grants.
Read visual as text
  1. Repository
  2. Files
  3. Terminal
  4. Network
  5. Secrets
  6. Database
  7. Pull requests
  8. Deployment

Default deny. Minimum capability. Explicit approval. Expiry and revocation.

INDEPENDENT VERIFICATION

An agent should not approve its own result

Planned verification checks the approved plan, exact repository state, diff, tests, build, artefacts, and delivery outcome outside the coding-agent session.

PLANNED

Completion is a claim until it is verified

01 · AGENT REPORTAgent-reported completionCompletion claim
02 · INDEPENDENTIndependent checksPlan · diff · tests · build
03 · EVIDENCEDVerified resultUpdated project state
Agent-reported completion is checked independently against the approved plan, exact diff, tests, build, artefacts, and delivery state.
Read visual as text
  1. Approved plan
  2. Agent-reported completion
  3. Independent checks
  4. Verified result
  5. Updated project state

Agent report ≠ verified result. Evidence closes the loop.

MVP v1.0 · LIVE

The live planning and project-intelligence foundation

Airchor currently turns user-provided context into persistent, readiness-labelled plans, assessments, roadmaps, and handoff drafts through two live routes.

Explore the current product
LIVE FOUNDATION

The live project-intelligence foundation

ROUTE 01Create New Project
ROUTE 02Work on Existing Project
Project context
  • Evidence
  • Readiness
  • Roadmaps
  • Handoffs
Create New Project and Work on Existing Project feed persistent context, evidence, readiness, roadmaps, and handoff documents.
Read visual as text
  1. Create New Project
  2. Work on Existing Project
  3. Project context
  4. Evidence
  5. Readiness
  6. Roadmaps
  7. Handoffs

This is live MVP v1.0 functionality. It does not execute or verify code.

ROUTE 01LIVE

Create New Project

Define a new software concept, boundary, assumptions, roadmap, and first handoff.

Output package

A deterministic 14-document Project Blueprint.

Explore Create New Project
ROUTE 02LIVE

Work on Existing Project

Organise supplied project and repository context for manual assessment and continuation planning.

Output package

A deterministic 10-document Assessment & Recovery Plan.

Current boundary: Airchor does not currently inspect or verify repository code automatically.

Explore Work on Existing Project

AUDIENCE ROUTES

One control-plane direction. Different decision boundaries.

FOUNDATION

Founders

Carry a defined problem, scope, assumptions, and acceptance criteria from project foundation toward governed implementation.

Explore route
ASSESSMENT

Agencies

Move from supplied client evidence and bounded findings toward a reviewable recovery direction and future governed repair.

Explore route
DECISION

Advisors

Keep evidence, uncertainty, options, risk, and the owner’s next decision visible around future execution.

Explore route
CONTEXT

AI Builders

Turn fragmented prompts and fast-moving implementation context into durable project state and controlled agent work.

Explore route
BOUNDARY

Freelancers

Define a responsible scope, explicit exclusions, acceptance criteria, and review points before another sprint.

Explore route
GATE

Agency Assessment

Create a deliberate evidence and decision gate before estimating or committing to implementation.

Explore route

PLATFORM ARCHITECTURE

Nine layers from project intelligence to visible execution

The planned Airchor platform separates planning, authority, execution, verification, evidence, integrations, and user control into distinct responsibilities.

PLANNED ARCHITECTURE

Nine control-plane layers

  1. Project intelligence
  2. Planning & decisions
  3. Agent orchestration
  4. Capabilities & permissions
  5. Execution
  6. Verification
  7. Evidence & audit
  8. Integration adapters
  9. User execution console
Nine planned layers separate project intelligence, planning, orchestration, permission, execution, verification, evidence, integrations, and the user console.
Read visual as text
  1. Project intelligence
  2. Planning & decisions
  3. Agent orchestration
  4. Capabilities & permissions
  5. Execution
  6. Verification
  7. Evidence & audit
  8. Integration adapters
  9. User execution console

A provider-neutral control plane coordinates specialist systems without replacing them.

Explore the Agentic VisionView the Roadmap

ROADMAP PROGRESSION

Build the control plane in dependency order

The roadmap separates live, active-development, planned, and vision states. It does not publish delivery dates or percentage-complete indicators.

ROADMAP

A dependency-led platform progression

  1. LIVE
  2. ACTIVE DEVELOPMENT
  3. PLANNED
  4. VISION
The live planning foundation supports active public-truth and repository-intelligence work before planned execution and the longer-range control-plane vision.
Read visual as text
  1. LIVE
  2. ACTIVE DEVELOPMENT
  3. PLANNED
  4. VISION

Sequence communicates dependency, not dates or percentage complete.

View the RoadmapExplore the Agentic Vision
CURRENT LIMITATION

Structure today. Governed execution is the next phase.

Airchor MVP v1.0 is a planning and assessment product. The current product prepares clearer project context and handoff drafts.

  • Planning and assessmentLive today
  • Automatic source-code inspectionNot current MVP functionality
  • Code execution or automatic repairNot current MVP functionality
  • User-project test or build executionNot current MVP functionality
  • Pull-request creationNot current MVP functionality
  • Project deploymentNot current MVP functionality
  • Technical, security, or production-readiness certificationNot current MVP functionality

WHY AIRCHOR EXISTS

Software can now be created faster than it can be understood

AI-assisted development has accelerated implementation, but context, evidence, decisions, boundaries, and handoffs often remain fragmented. Airchor exists to make that project state visible and govern what happens next.

Why Airchor exists
CONTROL THE NEXT ACTION

Explore the platform direction and its live foundation

Understand the Product, Agentic Vision, Roadmap, or open Airchor MVP v1.0.

Agentic VisionProductRoadmapOpen Airchorapp.airchor.com

PRODUCT UPDATES

Follow the next stage of Airchor

Receive occasional product, roadmap, and release updates from Airchor.

Subscription is separate from application access. You can unsubscribe from future updates.

Signup uses double opt-in. Airchor does not store your email in a local website database. Read the Privacy Policy.