skills
the 18 yf-* skills, grouped by what they do
yoshiko-flow ships 18 skills, grouped here by what they do — the workflows you invoke to get work done, the beads support layer they build on, and the beads-free utility and markdown helpers. User-invocable skills are triggered with /yf-<skill>; auto skills fire from their description conditions when relevant work appears. Install them all with yf skills install, or by install group with yf skills install --group <beads|utility|markdown> (the workflow skills ship as skill-group: beads since they need bd; see install). The Depends on column shows each skill's depends-on-skill closure — naming a skill at install time pulls its dependencies in automatically.
workflows
End-to-end, beads-tracked user workflows — the skills you invoke to get work done.
| Skill | Invocation | Purpose | Depends on |
|---|---|---|---|
yf-incubator | /yf-incubator | Create, fork, bookmark, resume, and triage research topics ("incubators") under Incubator/. | yf-beads-extra |
yf-plan | /yf-plan | Structured planning with beads-tracked execution and upstream issue reconciliation. | yf-beads-extra, yf-beads-authoring |
yf-research | /yf-research | Multi-phase, beads-tracked deep research: decomposes a topic into a DAG of focused subtasks (retrieve → triangulate → synthesize → critique → refine → package) and produces a structured, citation-backed report with source credibility scoring. | yf-beads-extra, yf-beads-authoring |
beads
The bd (beads) support layer the workflows build on: init/health, direct-CLI gotchas, authoring conventions, graph hygiene, and upstream tracking.
| Skill | Invocation | Purpose | Depends on |
|---|---|---|---|
yf-beads-authoring | auto | Conventions for building Claude Code skills that orchestrate work through beads (bd): formula authoring (.formula.toml), `bd mol pour` lifecycle, dynamic fan-out, agent metadata wiring, the coordinator dispatch loop, the coordinator resilience contract (crash/resume recovery, stuck-bead sweep, completion handoff), and the `coordinate` subcommand with gate auto-detection. | yf-beads-extra |
yf-beads-extra | auto | Advanced/gotcha layer for using the `bd` (beads) CLI directly at runtime, on top of the canonical `beads` skill. Covers issue-type semantics, dependency-edge mutation, gate semantics, defensive JSON parsing, transactional bulk intake (`bd batch`), and `bd mol pour` output shape. | — |
yf-beads-hygiene | /yf-beads-hygiene | Safe, read-only-first audit and gated repair of a beads (`bd`) dependency graph: finds orphaned beads and dangling dependency edges, and correctly classifies gate-typed edges so live gates are never mistaken for dangling. The canonical trigger for any "clean up beads" request. | yf-beads-extra, yf-beads-init |
yf-beads-init | /yf-beads-init | Verify, initialize, and repair a functioning beads (`bd`) configuration in a repository, and the shared dependency-verification home that other beads skills' preflights route to. | yf-beads-extra |
yf-beads-upstream | /yf-beads-upstream | Configurable, GitHub-first upstream-tracking skill for beads. Pushes open/deferred beads to an issue tracker (GitHub/GitLab/Jira) as a land-the-plane step, and enumerates upstream issues as the authoritative worklist on status/pull. | yf-beads-extra |
utility
Beads-free helper skills — no bd binary required.
| Skill | Invocation | Purpose | Depends on |
|---|---|---|---|
yf-change-validation | /yf-change-validation | Runs a repo's recorded VALIDATION RECIPE (build/test/lint) over a change-set or merged tree by EXECUTING the commands and reporting PASS / FAIL / INCONCLUSIVE + the first failing command; never auto-fixes and never auto-rewrites the manifest. Driven by a per-repo CHANGE-VALIDATION.md inferred from the toolchain, operator-approved, then re-proposed on drift. | — |
yf-diagram-authoring | /yf-diagram-authoring | Generate light-mode, white-background diagram PNGs from d2 source, keeping the .d2 source beside every .png render. Standardizes d2 (not mermaid) as the single, local, offline diagram engine: write .d2 -> render .png (theme 0, elk) -> verify by Read. | — |
yf-drift-check | auto | Verifies CONTENT AGREEMENT across a repository's declared source-of-truth edges (implementation ↔ docs ↔ spec); never authors, optimizes, restructures, or auto-fixes. On edit of a file matching the repo's DRIFT-CHECK.md manifest globs, dispatches an isolated, report-only sub-agent that checks each scoped edge under a strict evidence standard and returns PASS / FAIL / INCONCLUSIVE / CONFLICT. | — |
yf-okf | /yf-okf | Constructs, manages, and conformance-checks the OKF-compatible artifact folders ('bundles') that yf artifact-producing skills emit (yf-plan, yf-research, yf-incubator), and owns the versioned OKF-* spec family (BASELINE + YF-EXTENSIONS + per-skill OKF-EXTENSION). The engine (okf.py) composes the effective ruleset and runs check (report-only conformance) and migrate (opt-in, per-folder, in-place). | — |
yf-optimal-instructions | auto | Auto-fix skill for project instruction files. On create/modify of a project CLAUDE.md, AGENTS.md, AGENTS/*, or repo-root .{claude,agents}/rules/* file, reads it, auto-applies token-efficiency cuts, and proposes structural fixes — AGENTS.md primary, CLAUDE.md a thin @-include index, behavioral rules in the project rules surface — then reports what changed. | yf-skill-authoring |
yf-skill-authoring | auto | Conventions for authoring Claude Code skills, agents, and instruction files. Covers directory layout, the inline-vs-script threshold, modularization, token-efficient writing rules, AND Python helper scripts for skills (uv invocation discipline, PEP 723 inline deps, argument parsers). | — |
markdown
Standalone GitHub-Flavored-Markdown tooling, beads-free.
| Skill | Invocation | Purpose | Depends on |
|---|---|---|---|
yf-markdown-format | auto | The autofix side of `yf-markdown-lint` — rewrites Markdown in place to conform to plain GFM along the axes the linter flags. Owns two transforms: strict GFM table alignment (`--check` gate / `--write` idempotent autofix / bare stdout) and Obsidian to GFM wiki-link migration. | — |
yf-markdown-html | auto | Render a Markdown file to a single, self-contained HTML file via pandoc — standalone document, all resources embedded (images, CSS, fonts), a broad-coverage default stylesheet, relative image paths (``) resolved against the source file's directory, self-contained math (MathML, no CDN), and opt-in CriticMarkup rendering. | — |
yf-markdown-lint | auto | Conventional GitHub-Flavored-Markdown linter. Checks that documents are valid GFM with well-formed, resolvable links — no Obsidian wiki-links (`[[...]]`) or embeds (`![[...]]`), valid relative links/anchors, and consistent tables. | — |
yf-markdown-pdf | auto | Render a Markdown file to PDF via the pandoc + xelatex pipeline — xelatex engine, a broad-coverage Unicode font (so →, ≤, ≈ and similar glyphs render), 1in margins, blue links, and relative image paths (``) resolved against the source file's directory. | — |