Habitat Maturity Model (14 dimensions)¶
The Agentic Experience 5-Level Habitat Maturity Model is the spine of the assessment. It describes what a team's habitat actually delivers across fourteen dimensions, each placed L1–L5. Verbs are in bold — the verb is the finding.
| Dimension | L1 | L2 | L3 | L4 | L5 |
|---|---|---|---|---|---|
| Agent behaviour | Dictating | Commanding (prompting) | Regulating | Orchestrating | Supervising |
| Agent input | short ad-hoc prompts | larger prompts, commands | plans co-authored with an agent | iteratively refined specs | refined specs + customer/observable metrics |
| Workflow | safe runtime, generic | prompts/commands saved | harness engineered | workflow defined | workflow automated (agentic runtime) |
| Operating model | Chat with agent | Prompt-engineering | humans drive / verify | humans in the loop | humans certify |
| Teams provide | — | basic team-specific constitution | comprehensive product-specific constitution | full product-specific constitution | custom product-specific runtime |
| Output role (I am…) | Running | Inspecting | Standardising | Specifying | Certifying |
| Output artefact | executable / artifact | code | process & consistency rules | clear criteria | evidence |
| Humans review | output only | code | implementation in detail | specs | comprehensive evidence |
| Work patterns | partial task completion | small task completion | e2e development | semi-autonomous work | mostly-autonomous |
| Agent composition | single | single + saved patterns | primary + read-only critics | bounded ensemble (harness-composed) | self-orchestrating constellations |
| Agents… | Assist individuals | Complete basic tasks | Develop small changes (stories) | Implement larger changes (epics) | Implement autonomously |
| Testing | Manual inspection | Asserting (unit tests) | Verifying (functional / business) | Validating (comprehensive automation) | Assuring (multi-perspective + post-deploy) |
| Observability | Eyeballs | Captured | Instrumented | Aggregated | Closed loop |
| Governance | trust-based, ambient | conventional | Constitutional | Policy-as-code | Continuous certification |
The dimensions are scored L1–L5, not L0–L5: L1 is the "ad-hoc but present" floor. A repo with essentially no AI-collaboration evidence sits at the L1 floor on every dimension by definition.
The headline Habitat Maturity Level¶
The overall Habitat Maturity Level is the rounded mean of the fourteen placements, with the weakest dimensions named as the ceiling — a habitat is only as mature as the dimensions its work actually flows through. A high mean dragged down by one L1 dimension is reported as "L3, held back by L1 Observability", not a flat L3.
How each dimension is placed¶
Eight dimensions are repo-observable — placed evidence-first from the scan:
| Dimension | Signals that raise the placement |
|---|---|
| Workflow | saved prompts/commands (L2); a harness document — HARNESS.md / CONSTRAINTS.md (L3); defined workflow scripts or CI pipelines (L4); automated agentic runtime (L5) |
| Teams provide | richness of CLAUDE.md / AGENTS.md (basic → comprehensive → full constitution); product-specific skills; custom runtime / prod-like agent environments (L5) |
| Agent input | ad-hoc prompt traces (L1); saved prompt/command libraries (L2); plan documents (L3); a specs/ directory (L4); specs + observable metrics (L5) |
| Output artefact | raw artifacts (L1); code (L2); process & consistency rule docs (L3); acceptance-criteria documents (L4); evidence artefacts — audits, CI evidence (L5) |
| Agent composition | custom agents; read-only critic/reviewer agents; an orchestrator with safety gates; agent-team docs; multi-agent workflow scripts |
| Testing | test suites; coverage enforcement; mutation testing; tests-before-merge gates; system/regression suites; agent-authored tests; prod-like environments |
| Observability | agent-activity logging; metrics capture; dashboards; per-PR acceptance / mutation-kill tracking; OTel config; closed-loop signals |
| Governance | HARNESS.md constraint count + enforcement ratio; policy-as-code CI checks; the unverified → agent → deterministic promotion ladder; governance-audit cadence |
Six dimensions are behavioural — they describe how the team works, not what the filesystem holds. They're inferred from the repo-observable dimensions and from the clarifying questions, and flagged (inferred) when not directly evidenced: Agent behaviour, Operating model, Output role, Humans review, Work patterns, Agents….
Evidence-first vs survey¶
- Evidence-first (default) — repo-observable dimensions placed from the scan; behavioural ones inferred.
- Survey (opt-in) — marker statements administered on a 1–5 scale, two per level, for a rigorous per-dimension score. See Run the precise survey.
The four headline axes¶
Four dimensions — Agent composition (reported as Composition), Testing, Observability, Governance — are the most repo-observable and map cleanly onto the three disciplines. They're surfaced as the Operational Axes (Part D) table in the report. The Habitat Build Gap, however, uses the mean of all fourteen dimensions.
Provenance¶
The fourteen dimensions and their verbs are the Agentic Experience 5-Level Habitat Maturity Model (TechTalk.AI / Agentic Engineering). The AI Literacy framework's ALCI drew its four operational axes from this model; this instrument scores against the model in full. See The Sovereign Engineer.