Command & skill¶
The plugin ships two pairs of entry points — one for assessing a single subject, one for rolling several up. Each pair is a command and a skill carrying identical content, so either half produces the same result.
| I want to… | Invoke |
|---|---|
| Assess this repository | /ai-readiness-assess |
| Assess every subject in a scope, one session | /ai-readiness-assess --scope |
| Summarise assessments that already exist across repos | /ai-readiness-rollup |
The last two need a scope manifest; with none
present, /ai-readiness-assess behaves exactly as it always has.
The command¶
/ai-readiness-assess — the canonical entry point. Defined in
commands/ai-readiness-assess.md.
Scope mode¶
With a .habitat/scope.yml present, --scope assesses every subject in
one session, asking the behavioural questions once for the team
rather than once per repository:
Each subject gets its own report in its own assessments/ directory,
and a portfolio report is produced at the end. See
assess a team across repositories.
The roll-up command¶
/ai-readiness-rollup — summarises assessments that already exist,
across several subjects. Defined in commands/ai-readiness-rollup.md.
It runs no scan and asks no questions: it reads the summary block from reports that are already written. That is the recommended path for a large estate, and it separates who runs an assessment from who reads the summary. See roll up existing assessments and the portfolio report structure.
The skill¶
ai-readiness-assessment — the natural-language surface. Defined in
skills/ai-readiness-assessment/SKILL.md. It triggers on phrases like:
assess our AI readiness · run an AI readiness assessment · check our AI literacy level · where are we on the framework? · what level are we at? · score our AI maturity · how ready is this codebase for AI collaboration? · check our habitat maturity · evaluate how we work with AI
It does not trigger on general framework questions ("what is the Sovereign Engineer about?") — those are explanation, not assessment.
The roll-up skill¶
ai-readiness-rollup — the natural-language surface for the roll-up.
Defined in skills/ai-readiness-rollup/SKILL.md. It triggers on phrases
like:
roll up our assessments · assess across repos · portfolio view · how do our repos compare · multi-repo readiness · compare our teams' habitats
Dual-surface sync¶
The framework content embedded in each command and its skill must stay identical — editing one without the other is forbidden (it's a harness constraint on this repo). This is why the same result is produced whichever way you invoke it.
Both pairs are checked by the test suite rather than by discipline: the assertions compare the two bodies directly and name the first line that diverges.
Layout¶
ai-readiness-assessment/
├── .claude-plugin/
│ ├── plugin.json # plugin manifest
│ └── marketplace.json # techtalkai marketplace manifest
├── commands/
│ ├── ai-readiness-assess.md
│ └── ai-readiness-rollup.md
├── skills/
│ ├── ai-readiness-assessment/
│ │ └── SKILL.md
│ └── ai-readiness-rollup/
│ └── SKILL.md
├── specs/ # intent, first-class
├── tests/ # TDAB structural test suite (contributors only)
├── CHANGELOG.md
└── HARNESS.md # this repo's own harness
Process the instrument follows¶
- Scan — habitat-document discovery, then a broader signal scan.
- Present & question — a structured summary, then 3–5 clarifying questions one at a time.
- Assess — the fourteen-dimension profile, the cognitive read, and the Habitat/Workflow Gap.
- Document — write the report.
- Reading path & Next steps — gap-anchored book chapter and a single TechTalk engagement.
- Offer the HTML render on request.
In scope mode the behavioural questions in step 2 are asked once for the team; each subject is then asked only whether its way of working differs. A reused cognitive read is always declared in the report body.