CHAPTER 01
Operating thesis
Buying AI coding tools gives individual engineers an engine. Apptension installs the factory around those engines: shared context, task standards, skills, verification, PR evidence, autonomy boundaries, telemetry, coaching, and an improvement loop.
Category we operate in
Agentic Enablement / AI-Native SDLC. The service is operating change with measured engineering outcomes, not tool training, prompt libraries, licence resale, or a promise to replace developers.
What we install
- 01Advisory and baseline assessment before claiming uplift
- 02Repository and context engineering agents can actually use
- 03Agent-ready planning standards: Intent → Constraints → Context → Acceptance → Verification
- 04Reusable skills and workflows wired into delivery
- 05Automated testing, evals, and CI gates
- 06Evidence-rich PR workflows
- 07Autonomy and approval policies matched to risk
- 08Engineering telemetry and economics
- 09Developer coaching and continuous harness maintenance
What we do not sell
- 01Cursor, Claude, or Copilot training alone
- 02Prompt libraries without delivery integration
- 03AI licences
- 04Developer replacement
- 05Unbounded autonomous production deployment
- 06Productivity claims from LOC, commits, prompts, or tokens
Operating principles
Sell operating change, not tool adoption
Establish baseline before claiming uplift
Treat quality, security, and human control as constraints on productivity
Prefer client-owned code, cloud, context, rules, telemetry, and skills
Use a bounded 2-week install for momentum, then managed adoption so the harness does not drift