interview-me

Reference for writing any document an agent consumes: a skill's SKILL.md, an AGENTS.md/CLAUDE.md, or a doc reached by a pointer from one.

Shared vocabulary for designing deep modules: module, interface, depth, seam, adapter, leverage, locality.

Active discipline for a project's visual/interaction rules: design tokens, component conventions, and interaction rules, plus an optional scratch doc for…

One-time setup of the project-memory system in a new project: creates the config file, the budgeted memory file, the memory README, and (optionally) a…

Active glossary/ADR-lite discipline for a project's domain terms: define terms, challenge fuzzy language before it spreads through the codebase, stress-test…

Shared interview primitive: model open decisions as a tree, work it in rounds via CoCo's ask_user_question tool, and iterate until nothing is left silently…

Explain how the project-memory system works, either generically (if this project isn't set up yet) or concretely using this project's actual config.json and…

A relentless interview to sharpen a plan or design, which also routes resolved decisions into the project's domain-vocabulary and/or design-soul docs as…

Grep-based, config-gated, opt-in drift checker: banned synonyms used as identifiers, documented code references that no longer exist, and multi-file value…

Run at the start of every session in a project that uses the project-memory system.

End-of-session wrap-up for projects using the project-memory system.

Audit an already-bootstrapped project-memory setup for drift: a stale configVersion against the installed plugin, dangling canonicalDocs/ownershipDoc paths…