claudialnathan

state-bump

Refresh STATE.md against the live Claude Code changelog. Edits the "Last updated:" / "Version:" header on line 3, then adds, edits, or retires body entries so the snapshot matches what's at code.claude.com/docs/en/changelog. Preserves the document's existing voice (terse, factual, "what changes how you build, configure, and ship"), its section structure, and the dateline convention, which means header fields only and no inline "last reviewed against vX.Y.Z" reintroduced elsewhere. On every substantive bump, also re-pin and amend HACKS.md, check CLAUDE.md for staleness, and sweep the kitchen's skills/ artifacts against the new state (Steps 5–7).

forge

Designs Claude Code harness artifacts. One triage ladder picks the right surface (skill, hook, path-scoped rule, CLAUDE.md or AGENTS.md entry, workflow script, subagent, MCP); the stance holds the bar every artifact must clear (a gradeable objective, a gap between the model's default and what the owner wants, and a nameable attention shift; the gap is a new idea or source worth operationalizing, a miscommunication to prevent, or an observed failure); per-surface mechanics live in references. Use for creating, reviewing, or refactoring any of these, for distilling articles or docs into one, for skills that under- or over-trigger, hooks that don't fire or fire too often, bloated CLAUDE.mds, and for deciding which surface a behavior belongs on. Triggers on any work on a skill, hook, rule, CLAUDE.md, AGENTS.md, or workflow orchestration script, including debugging triggering or firing semantics and packaging a /command or plugin.

harvest

Mines Claude Code session transcripts for the corrections that earn artifacts, the intake stage of the kitchen's loop. Extracts what the user actually said across recent sessions (via a bundled script that reduces multi-MB transcripts to user turns), dispatches readers to classify corrections, redirects, and repeated failures, dedups against the standing harness, and hands surviving candidates to the forge with evidence counts. Use on a cadence (end of a heavy week, after a project push), when a correction feels familiar, or before claiming a harness has no gaps; skip when recent sessions were light or single-topic.

ingest

Grounds the next forge step in supplied reading, weighing sources the way the owner handed them. Material the request is built on (inspired-by, do-what-they-do, this person's articles, "the primary resources") is read whole in the main thread; supplementary references fan out to parallel subagents (one per source) that return quoted excerpts. Both land in a bounded brief with citations that hands off to the forge skill with the right surface in mind. Use when you have reading material (URLs, papers, blog posts, internal docs, pasted content) and want the next forge step grounded in those sources rather than training-data priors; skip when training already covers the topic well.

harness-audit

Audits a Claude Code setup end-to-end: what loads every session, what it costs, and whether the harness's own claims hold. Inventories the standing surfaces (CLAUDE.md chain, personal/project/plugin skills, MCP instruction blocks, agent rosters, hooks, settings at every scope), quantifies their per-session token cost, runs five checks (self-consistency, duplication across scopes, enforcement parity, scope discipline, and a provisional routing-effectiveness check for doc-fleet repos) and reads the CLAUDE.md chain as intent (code outranks the doc; no restated stack, deps, or layout). Output is a quantified triage that stays in-repo: findings at machine scope (anything under ~/.claude, user or enterprise settings, global plugins) are reported for the owner to action, never edited. Use to review, slim, or sanity-check a whole Claude Code setup (including 'why is my context so big' or 'what loads every session') after major harness changes (a new plugin fleet, a new machine, a model release) or when sessions feel