kunalsuri

Bootstrap the ai/ knowledge layer for a repo that has none — draft the MODULE_MAP, diagrams, and guide docs as [inferred] for a human to audit, touching no source code. Use when the maps are still placeholders or the user asks to cold-start, bootstrap, or onboard a repository.

Bootstrap the ai/ knowledge layer for a repo that has none — draft the MODULE_MAP, diagrams, and guide docs as [inferred] for a human to audit, touching no source code. Use when the maps are still placeholders or the user asks to cold-start, bootstrap, or onboard a repository.

Deep, judgement-based adversarial code audit — hunts for stale cross-references, unescaped interpolation, platform gaps, generated-file ownership conflicts, and cross-module consistency rot that no mechanical check can catch. Read-only; writes a dated findings report.

Kit-maintainer diagnostic (LOCAL-ONLY, not distributed) that audits the README and docs/ for staleness against the repo's real command/skill/CLI roster, stamped tree, and cross-doc consistency. Read-only; produces a severity-ranked findings report.

Run verification and drift checks on the AI knowledge-base to identify missing documentation or stale references.

Deep-mine the source code to discover implemented features; writes ai/analysis/FEATURE_CATALOG.md.

Implement a spec from ai/lab/specs/ exactly as written — zero design decisions, stop-and-report on any spec-vs-reality conflict, tests are the definition of done. Built for lighter implementation models.

Dry-run the add-feature workflow for a proposed feature — friction report and per-phase readiness score, without writing code.

Deep, judgement-based adversarial code audit — hunts for stale cross-references, unescaped interpolation, platform gaps, generated-file ownership conflicts, and cross-module consistency rot that no mechanical check can catch. Read-only; writes a dated findings report.

Audit every ai/ file for gaps, stale placeholders, and inconsistencies after cold-start. Produces a prioritized findings report.

Diagnostic gate that checks AGENTS.md and AGENTS.md for completeness, consistency, and stale artifacts — run right after /cold-start and before /add-feature. Read-only; produces a findings report.

Typecheck and run the full test suite, and report failures concisely

Holistic assessment of the AI knowledge layer on a 5-level maturity scale; flags agent-blocking gaps.

Creates or refreshes CLAUDE.md and AGENTS.md at a repo's root — both carrying the same real project context (stack, build/test/lint commands, structure, conventions) — plus scaffolds the accompanying .claude/ and .agents/ project directories, so Claude Code, Antigravity, Cursor, GitHub Copilot, Codex, and other AI coding tools all start from identical instructions instead of silently drifting out of sync. Use this whenever the user wants to set up a repo for multiple AI coding assistants, asks for both a CLAUDE.md and an AGENTS.md (or just one but mentions wanting the project to also work in another tool), wants a project-level .agents/ or .claude/ folder created, or says things like "make this repo work in Claude Code and Antigravity", "add agent instructions for VS Code", "this project needs to work across coding assistants", or "sync my AI context files". Also trigger if a repo already has a CLAUDE.md or AGENTS.md that looks stale, contradicts the other, or was never kept in sync.

Runs a multi-step implementation task as a resumable, git-committed loop -- write a spec doc, derive a checklist from it, then implement/test/commit/check-off one item at a time. A lightweight take on spec-driven development. Use when a task is large enough to span multiple sessions, could be interrupted mid-way, or when the user asks for spec-driven development, SDD, or a process that survives the agent stopping unexpectedly (session death, context overflow, running out of budget) so no work is lost.

Add a feature to this repository the safe way — spec, locate via maps, respect Stability, surgical implementation, verification, knowledge update. Use whenever the user asks to add, build, or implement functionality.

Run the SkillDeck kitchen pipeline locally, doing capability clustering and card writing yourself instead of downloading an embedding model or calling an LLM API.