---
id: roadmap
title: Roadmap
description: Build proven skills in human and AI Agent work, starting with Delivery, as the foundation for the future autonomous SDLC.
hide_table_of_contents: true
---

import ArchifyDiagram from '@site/src/components/ArchifyDiagram';

<div className="strategy-content">

## Two stages toward Target

<ArchifyDiagram name="strategy-roadmap" title="Adopt skills → Future Autonomous Agent" story />

**Humans direct, guide and review throughout.** Stage 1 standardizes skills in human + AI Agent work across the SDLC. Proven skills, measurements and working practices become the foundation for Stage 2: the future Autonomous Agent in [Target](./target-operating-model.mdx).

Target operating model—not deployed autonomy, an agreed schedule or approved KPI thresholds.

## Stage 1: Delivery first

<ArchifyDiagram name="strategy-roadmap-adoption" title="Delivery first → Other SDLC phases, order undecided" story />

Start with Delivery, then extend skills adoption to the remaining SDLC phases. Their order is not yet decided. **Delivery keeps improving as adoption expands.**

## Improve within each phase

<ArchifyDiagram name="strategy-roadmap-loop" title="Each phase: understand → baseline → topology → use → collect → reassess → repeat" story />

Repeat this loop within every phase—not as six global roadmap stages. Skills topology defines **what skills exist, how they relate, when they are used and how they work together**, including handoffs—not just a catalogue.

## Delivery in practice

**Understand the engineering path first:** codebase structure and existing principles → system-logic source → tests → builds and packaging → deployments. Inspect actual practices, tooling, controls and problems with the people doing the work, rather than assuming the documented process is the real one.

**Establish the comparison, then engineer skills.** Define metrics/KPIs and a baseline for comparable work before changing it. [Skills Measurement](./measurement.mdx) explains collection and calculation. Connect reviewed skills to the engineering path, with clear triggers, dependencies and acceptance checks.

**Use, collect, compare, repeat.** A human works with an AI Agent on real changes. Collect results, friction and feedback alongside quality, cost and verification effort. Compare with the baseline to understand the new current situation; revise the skills and working practices, then run the loop again.

<details>
<summary>Supporting gate evidence and hold conditions</summary>

These are recommended implementation gates, not assigned human owners, approved thresholds or installed integrations. Agree acceptance criteria for each selected workflow before evaluating it.

### Understand the current situation

Sample representative repositories and changes. Retain repository revisions, tests and CI reliability, maintainability, technical debt and security findings. Trace ownership, review and release practices, handoffs, rework and bottlenecks. Keep baseline findings, a process map and prioritized gaps; hold selection if the sample cannot explain how work is delivered and accepted.

### Prepare safe, measurable human + AI work

Inventory actual client hosts/builds, provider/data terms, authority and event availability. Confirm accountable people and independent reviewers, forbidden actions, data/retention decisions, recovery and a reproducible baseline. Hold if accountability, lawful collection or critical acceptance checks remain unresolved.

Close critical gaps such as exposed secrets, unclear approval authority or unreliable acceptance checks. Broader refactoring and non-critical debt can run alongside the selected work. Stage 1 introduces no new privileged autonomous actions.

### Engineer and evaluate connected skills

Use a small registry, immutable skill contracts, qualified client adapters, evaluations and coaching. Verify the measurement calculations with synthetic fixtures before interpreting real outcomes. Require mandatory safety checks, independent task acceptance, reconciled source coverage and user feedback. Hold or narrow scope if quality or coverage is inadequate; usage alone does not justify expansion.

### Extend adoption without stopping improvement

Prove canary updates, effective-version readback, version locking, exception expiry, stale/offline handling, rollback and revocation. Hold if effective versions or controls cannot be verified or the catalog cannot be supported.

Use evidence to choose the next SDLC phase; no second or third phase is assigned here. Continue the improvement loop in Delivery and every phase already using skills.

### Prepare for the future Autonomous Agent

Stage 2 builds on proven skills, measurements and working practices—not permission inferred from adoption alone. Before granting scoped execution authority, prove observation and evidence assembly, then proposing and routing work. Recommended controls include exact-artifact approvals, idempotency, duplicate/outage recovery, separation of duties, audit and kill-switch drills. Prove human takeover; never infer permission from missing evidence. The agent ultimately executes and governs the cycle, not only routes tasks, while humans retain direction and review. Keep sensitive decisions human. Changed client, provider, data or authority boundaries reopen the relevant checks.

</details>

**Next:** [Return to the strategy destination →](./target-operating-model.mdx)

</div>