ai-native-toolkit

Compare two versions of an LLM-directed document - an original (teacher) and a candidate (student) - across a transfer set and return a per-case behavioural-equivalence verdict plus an efficiency signal. A transform-agnostic library capability other skills compose to gate a transform on behavioural sameness. TRIGGER when asked to A/B test two versions of a prompt / instruction / skill, to check whether a rewritten or compressed document still behaves the same as the original, to validate behavioural equivalence between two document versions, to gate a transform on no-regression, or when a skill needs the run-the-runner-on-both-versions-and-judge-equivalence capability.

Assess a codebase's readiness for AI agent contributors using the layered contract model, and generate a complexity hotspot SVG treemap (size = LOC, hue = cyclomatic complexity, saturation = recent git churn). TRIGGER when the user types /assess, asks for an AI-readiness review, wants a complexity heatmap or hotspot map, asks 'how complex is this code?', wants migration risk triage, or asks for a codebase snapshot/report. Produces an MD report + SVG that can be opened as a PR in the target repo.

Renders the /assess report from the deterministic run-context.json and the layer scorecard - the scorecard, the verbatim cross-layer findings, lying signals, and the mandatory Top 3 Actions. TRIGGER when the /assess orchestrator reaches the report-writing step; not a standalone user command.

The /assess end-of-run offers - open a PR with the report, track the Top 3 Actions in the user's issue tracker, freeze the assessment into a CI gate, and file tool feedback. TRIGGER when the /assess orchestrator reaches the end-of-run offers; not a standalone user command.

Detect and remove the telltale signs of AI-generated 'slop' from any written text - articles, reports, emails, essays, bios, marketing copy, documentation, encyclopedia entries, or anything meant to read as if a thoughtful human wrote it. Apply silently as a quality gate before finalizing substantial prose, and explicitly when asked to clean a draft. TRIGGER when the user says 'make this sound less like AI', 'remove the AI tells', 'de-slop this', 'check if this reads as AI-written', 'make it sound human', 'edit out the ChatGPT voice', or critiques a draft as generic, puffy, or robotic. Based on Wikipedia's 'Signs of AI writing' field guide.

Read-only org repo state report. Reuses ghsync's repo discovery (teams union org-repo-list) but, instead of cloning, queries each repo's remote GitHub state - open PRs, CI on the default branch, open security alerts (Dependabot / code-scanning / secret-scanning), and branch protection - and rolls it into a terminal summary plus a timestamped markdown file. Endpoints that need admin and return 403/404 are reported as no-access / unknown, never as a clean zero. TRIGGER when the user types /ghreport, asks 'what state are the org's repos in', wants an org health report or repo state snapshot, or asks about open PRs / failing GitHub Actions / security alerts across a whole org or personal account.

Bulk clone and keep in sync every GitHub repo you can access across an org or personal account. For an org it unions the repos reachable through the teams you belong to with the org's repo list (so direct-collaborator and public repos count even with no team membership); for a personal account it uses the account's repo list. Deduplicates them, then clones new ones and fast-forward syncs existing ones (and their worktrees) into a consistent <repo>/<repo>-main + <repo>/worktree layout. The org defaults to the directory you run from, so dropping into ~/dev/github.com/<org> mirrors that org. TRIGGER when the user types /ghsync, wants to clone or sync all org repos or their own personal repos, asks to mirror everything they have access to, is onboarding into a new enterprise/org, or wants to refresh their local checkouts to latest.

Structured multi-perspective analysis using Six Thinking Hats with professional lens team members. TRIGGER when the user types /huddle, asks to run a huddle, wants a panel/board/team to analyze a decision, asks for multi-perspective analysis, debate, or red-team/blue-team review, or wants to weigh a hard call from several angles. Scales from solo (1 agent) to board-level (8+) using Fibonacci sizing.

Run a list of work units to completion with an Agent Team: derive a dependency DAG and hot-file map, spawn one ephemeral teammate per unit (or combined group), drive each PR through pr-review-merge, smart-merge in waves, recover from crashes, and run a retrospective. Source-agnostic — the caller supplies a work-source adapter. A library skill invoked BY the /tm and /issues commands, not run directly by a user (it needs a caller-supplied adapter). TRIGGER when a command needs autonomous multi-unit team orchestration to completion — a tag, issue queue, backlog, or set of tickets run to done with Agent Teams. For a single PR use pr-review-merge instead; not for one-off single-task work.

Drive a single pull request to merge-ready across all five criteria (sync, CI, inline comments, conversation, threads), then smart-merge it. Source-agnostic library skill invoked by the /tm, /issues, /fix-pr, and /fix-develop commands and by marathon teammates. TRIGGER when a command or agent needs the PR review-to-green loop or the smart-merge (stale-bot-CR dismissal, auto-merge criteria, UNSTABLE/UNKNOWN handling, merge ordering), or when the user asks to take a PR to green/merge it.

Make an LLM-directed document smaller while preserving what it does. Two modes: a local span-level core->pointer pass, and an A/B-validated distill loop that produces the smallest document that behaves the same as the original. Point at core knowledge the model already holds (a concept name activates it); keep project-specific detail explicit and verbatim. TRIGGER when asked to compress, tighten, shorten, or strip a prompt / instruction / system message meant for an LLM; to distill a skill; to compress a whole document; to make this smaller while preserving behaviour; to A/B test a compression or produce a behaviourally-equivalent compression; when an instruction set explains concepts the model already knows from training; or when reducing token cost of an LLM-directed prompt without losing meaning. Not for human-facing prose - that is /deslop.

Harden a skill or any agent instruction file (CLAUDE.md, AGENTS.md, GEMINI.md, .cursor/rules, .github/copilot-instructions.md) through judge-panel refinement rounds until it clears a 3-tier promotion gate, then promote it. A quality gate that runs after authoring, not an authoring tool. TRIGGER when the user types /skill-forge, asks to test/harden/forge/prove a skill or an instruction file, wants a CLAUDE.md or AGENTS.md driven through adversarial rounds before shipping, asks 'is this skill ready?' or 'is this CLAUDE.md safe for a cold-start agent?', or wants a skill or instruction file quality-gated by a judge panel.