SDLC blueprint

Handbook

Versonas — discipline challenge agents

**Versonas** are AI challenge agents that pressure-test work from discipline-specific perspectives. They are **not** team roles — they are challenge functions that strengthen thinking before costly co

Versonas — discipline challenge agents

Versonas are AI challenge agents that pressure-test work from discipline-specific perspectives. They are not team roles — they are challenge functions that strengthen thinking before costly commitments.

How Versonas work

  1. A decision point is reached (refinement, pre-build, pre-release, architecture choice).
  2. The team identifies which discipline perspectives are relevant.
  3. The appropriate Versonas are invoked (via Cursor rule or manual prompt).
  4. Each Versona challenges the work from its discipline's perspective, producing structured output.
  5. The team acts on concerns or accepts the risk, capturing the decision in the Ember Log.

Bridge-awareness

Each Versona template references its discipline's bridge document (*-SDLC-PDLC-BRIDGE.md). The bridge contains a phase alignment table that tells the Versona agent when its discipline is most relevant:

  • Architecture Versona is strongest at phases A–C (discover, specify, design), lighter at D–F.
  • Testing Versona activates at phases D–E (build, verify) and during Assay Gate.
  • BA Versona is strongest at phases A–B (discover, specify) and at Review.

This phase-awareness lets Versonas calibrate challenge intensity based on the current Spark's phase prefix (discover:, build:, verify:, etc.).

Template catalog

Engineering family (7 disciplines)

Template Discipline Core challenge
versona-se.mdc.template Software Engineering Are CS fundamentals and craft practices sound?
versona-architecture.mdc.template Software Architecture Is this structurally sound and maintainable?
versona-devops.mdc.template DevOps Can we deliver and operate this reliably?
versona-testing.mdc.template Testing & QA Can we prove this works correctly?
versona-frontend.mdc.template Frontend Is the web UI fast, accessible, and maintainable?
versona-mobile.mdc.template Mobile Is the mobile experience performant and reliable?
versona-iot.mdc.template Embedded / IoT Is this reliable and safe for constrained environments?

Data family (2 disciplines)

Template Discipline Core challenge
versona-bigdata.mdc.template Big Data & Data Engineering Is data engineered, governed, and processed correctly?
versona-datascience.mdc.template Data Science & ML Are models valid, reproducible, and responsible?

Product family (5 disciplines)

Template Discipline Core challenge
versona-product-management.mdc.template Product Management Are we building the right product for the right market?
versona-ba.mdc.template Business Analysis Do we understand what stakeholders need?
versona-ux.mdc.template UX / UI Design Is this usable, desirable, and accessible?
versona-marketing.mdc.template Marketing Can we acquire, engage, and retain users?
versona-cs.mdc.template Customer Success Will users achieve their goals?

Governance family (1 discipline)

Template Discipline Core challenge
versona-pm.mdc.template Project Management Are we delivering within constraints?

Cross-cutting (2 disciplines)

Template Discipline Core challenge
versona-security.mdc.template Security Is this safe from attacks and breaches?
versona-compliance.mdc.template Compliance Does this meet regulatory obligations?

Methodology / demo (1)

Template Discipline Core challenge
versona-sampling.mdc.template Sampling (demo) What assumptions, unknowns, and top signals appear before a full Versona pass?

Family aggregators

Template Activates
versona-family-engineering.mdc.template All 7 engineering Versonas
versona-family-data.mdc.template Both data Versonas
versona-family-product.mdc.template All 5 product Versonas
versona-all.mdc.template Master routing — suggests which Versonas based on context

Tasklets (example bundle)

Small single-operation Cursor rules and an installer script live in Forge tasklets — small Cursor operations. They demonstrate how a meta-Versona (e.g. Sampling) can chain tasklets before deeper discipline Versonas.

Discipline overlay (neutral tasklets + normative Versona)

flowchart LR T[Tasklet_mechanical_IO] V[Versona_discipline_overlay] R[Versona_contract_output] T --> V V --> R

Sampling Versona sequence (demo)

sequenceDiagram participant User participant S as Sampling_Versona participant A as Tasklet_assumptions participant U as Tasklet_unknowns participant Q as Tasklet_triage User->>S: Invoke S->>A: Charter_1 A-->>S: Output S->>U: Charter_2 U-->>S: Output S->>Q: Charter_3 Q-->>S: Output S-->>User: Merged_challenge

Where documentation is published

flowchart LR BP[blueprints_GitHub_MD] BPW[blueprints_handbook_HTML] FS[forgesdlc_com_HTML] BP -->|submodule| BPW BP -->|submodule| FS

Adopting Versonas in a consuming repo

  1. Copy the templates you need to .cursor/rules/ in your repo (remove .template suffix).
  2. Optional: Install the example tasklets + Sampling Versona via tasklets/install-tasklets.sh.
  3. Update globs: in each rule to match your project's file structure.
  4. Configure which Versonas are active in forge.config.yaml (via the setup wizard).
  5. Use family aggregators to activate discipline groups without configuring each individually.

See Versona contract for the standard structure every Versona rule must follow.

Canonical source

Edit https://github.com/autowww/blueprints/blob/main/sdlc/methodologies/forge/versona/README.md first; regenerate with docs/build-handbook.py.