mif-docs genre and CLI catalog
mif-docs genre and CLI catalog
Section titled “mif-docs genre and CLI catalog”An exhaustive lookup for the mif-docs v0.5.0 plugin: every skill it ships,
every doc-set recipe, and every script in scripts/ with its arguments and exit
behavior. Consult an entry; do not read this end to end.
Genre skills (38)
Section titled “Genre skills (38)”Each genre ships good-l1.md (L1 floor), good.md (target level), bad.md, and
evals/evals.json.
| Skill | Artifact it authors | MIF conceptType | Target level |
|---|---|---|---|
diataxis-tutorial |
Learning-oriented hands-on lesson | procedural |
2 |
diataxis-how-to |
Task-oriented recipe | procedural |
2 |
diataxis-reference |
Information-oriented lookup of one thing | semantic |
3 |
diataxis-explanation |
Understanding-oriented discussion | semantic |
3 |
arc42-arch-doc |
arc42 12-section architecture document | semantic |
3 |
c4-model-diagram |
C4 model document (Context/Container/Component/Code) | semantic |
3 |
google-design-doc |
Google-style engineering design doc | semantic |
3 |
engineering |
Engineering decision / evaluation report (options-vs-criteria comparison table) | semantic |
3 |
adr |
Architectural Decision Record (Structured MADR) | semantic |
3 |
rust-rfc |
Rust-style RFC / enhancement proposal | semantic |
3 |
python-pep |
Python Enhancement Proposal | semantic |
3 |
changelog |
Keep a Changelog 1.x version history | episodic |
2 |
sre-runbook |
Tactical step-by-step incident runbook | procedural |
2 |
playbook |
Strategic multi-incident operations playbook | procedural |
3 |
prd |
Product Requirements Document | semantic |
3 |
feature-spec |
Lightweight AI-ready feature specification | semantic |
2 |
ai-architecture-doc |
Composite AI-spec architecture document | semantic |
3 |
kiro-requirements |
Kiro requirements.md (user stories + EARS) |
semantic |
3 |
kiro-design |
Kiro design.md (technical design) |
semantic |
3 |
kiro-tasks |
Kiro tasks.md (checkbox implementation plan) |
procedural |
2 |
academic |
Peer-review-style IMRaD scholarly paper | semantic |
3 |
systematic-review |
PRISMA systematic review / meta-analysis | semantic |
3 |
computing-paper |
ACM/IEEE computing paper (IEEE numbered citations) | semantic |
3 |
humanities-mla |
MLA-style humanities paper | semantic |
3 |
humanities-chicago |
Chicago-style humanities paper (notes-bibliography) | semantic |
3 |
clinical-submission |
Clinical study report (ICH E3 CSR / CTD module frame) | semantic |
3 |
nist-sp |
NIST Special Publication structure | semantic |
3 |
regulatory-disclosure |
SEC-style annual disclosure (Reg S-K / Form 10-K item order) | semantic |
3 |
compliance-audit |
Compliance audit report (SOC 2-style controls + findings) | semantic |
3 |
security-pentest |
Penetration-test report (PTES/OWASP-style, CVSS findings) | semantic |
3 |
legal-memo |
Predictive legal memorandum (IRAC, Bluebook citations) | semantic |
3 |
market-research-report |
Market research report (methodology/fieldwork-grounded) | semantic |
3 |
business-plan |
Investor/lender-ready business plan (SBA/SCORE structure, Lean Canvas framing) | semantic |
3 |
sustainability-report |
GRI-Standards sustainability/ESG report | semantic |
3 |
trend-analysis |
Trajectory report (drivers/inhibitors, scenario diagram) | semantic |
3 |
competitive-quadrant |
Two-axis competitive-quadrant report (Mermaid quadrant chart) | semantic |
3 |
briefing |
One-page briefing / standup update | episodic |
3 |
exec-summary |
1-2 page BLUF decision-oriented executive summary | semantic |
3 |
Substrate skills (7) and the planner
Section titled “Substrate skills (7) and the planner”These do not author a finished genre artifact; they supply or check the MIF layer that every genre rides on, plus the multi-document engine.
| Skill | Role |
|---|---|
mif-frontmatter |
Author MIF Level 1–3 frontmatter for any document, climbing L1→L3 as detail allows. |
ears-acceptance-criteria |
Turn a requirement or driver into an EARS-notation acceptance criterion. |
mif-validate |
Deterministically prove a document is MIF-conformant and convert between forms. |
mif-corpus |
Semantically index and query MIF docs via the optional mif-rs tools; suggestion signal, never a gate. |
mif-provenance |
Stamp hook-witnessed provenance into frontmatter and verify blocks against the session ledger; witnessed, never asserted. |
svg-charts |
Generate a standalone .svg chart file (and its <img> embed line) for cases beyond Mermaid’s documented range. |
mif-to-pdf |
Convert a MIF JSON-LD document to PDF, embedding every frontmatter field as PDF metadata (Info dictionary + a custom lossless XMP packet). |
doc-set-planner |
Plan a subject into a coordinated SET of documents, fan out to genres, reconcile the relationship graph. |
Doc-set recipes (4)
Section titled “Doc-set recipes (4)”doc-set-planner decomposes a broad subject using one of four recipes. Each
recipe names its member genres and the cross-document relationship graph that
planner-check verifies for link-completeness.
| Recipe | Decomposes into |
|---|---|
diataxis |
tutorial + how-to + reference + explanation |
ai-spec |
AI-ready specification set (feature-spec + ai-architecture-doc + EARS criteria) |
kiro |
Kiro three-document set: requirements + design + tasks |
architecture |
Architecture set: arc42 + C4 + ADRs |
Scripts (scripts/)
Section titled “Scripts (scripts/)”All scripts run on Node.js 20+ and are invoked as node scripts/<name>.mjs.
Every script is fail-closed: any failure exits non-zero.
| Script | Invocation | Exit behavior |
|---|---|---|
mif-validate.mjs |
mif-validate <file> [--level 1|2|3] [--no-roundtrip] |
0 valid; 1 schema/level/round-trip failure; 2 usage error (no file); 3 schema cache not hydrated locally (environment gap, not a document failure — run npm run hydrate-schema). |
mif-convert.mjs |
mif-convert <emit-jsonld|emit-markdown|roundtrip> <file> [--no-check] |
0 success; 1 schema check / non-lossless round-trip; 2 usage error. |
mif-to-pdf.mjs |
mif-to-pdf <doc.json> [--output out.pdf] |
0 PDF written; 1 unreadable/unparseable input or missing required MIF L1 field; 2 usage error (no file). |
hydrate-schema.mjs |
hydrate-schema [latest|<version>] |
0 schema cached + VENDOR.lock written; 1 fetch failure (reports last hydrated version). |
hydrate-ontology.mjs |
hydrate-ontology |
0 ontology cached from published URI / ontologies repo / local sibling checkout; 1 unresolved. |
validate-ontology.mjs |
validate-ontology |
0 ontology valid and all entity_type / relationship types resolve; 1 not hydrated or dangling reference. |
validate-plugin.mjs |
validate-plugin [<plugin-root>] |
0 plugin.json, marketplace.json, .mcp.json, every SKILL.md frontmatter and its evals/evals.json, and every commands/**/*.md and agents/**/*.md frontmatter valid; 1 any structural violation. |
check-exemplars.mjs |
check-exemplars |
0 every genre’s good-l1.md validates at L1 and good.md at its target level; 1 otherwise. |
planner-check.mjs |
planner-check [<recipe>] |
0 recipe(s) decompose to real member skills and the cross-link graph is complete; 1 otherwise. |
mif-provenance.mjs |
mif-provenance <stamp|verify> <file> [--session <id>] [--ledger <path>] | mif-provenance status [--session <id>] [--ledger <path>] |
0 stamped / verify match / status healthy; 1 verify drift (including unwitnessed) or status found no session_start line for this session yet; 2 usage/environment error (unknown verb, no repo and no --ledger, ambiguous session, no session id available); 3 stamp declined (unwitnessed, not conformant, would regress the document’s MIF level). |
provenance-corpus-check.mjs |
provenance-corpus-check [--dir <path>] [--ledger <path>] |
0 report emitted (byte-idempotent over identical inputs — never gates on coverage); 1 empty corpus; 2 usage error. |
engine-parity.mjs |
engine-parity <path-to-mif-cli> [--expected <json>] |
0 node and Rust verdicts agree everywhere outside the expected-disagreement ledger; 1 unexpected disagreement, stale/orphaned ledger entry, or harness fault (missing/unparseable ledger, schema not hydrated, binary cannot run, gutted corpus); 2 usage error (missing binary path, wrong working directory). |
Which engine is authoritative
Section titled “Which engine is authoritative”The node engine is authoritative for every gate in this suite (ADR-0004):
mif-validate, the CI validation jobs, and the fail-closed write guard all
run scripts/lib/projection.mjs. The mif-rs Rust engine is compared against
it by the non-required nightly engine-parity workflow; a disagreement is
evidence for an upstream mif-rs issue, never a reason to change a document
the node gates accept.
Argument notes
Section titled “Argument notes”mif-validate --leveldefaults to1. The level overlay layers required fields on top of the canonical schema: L2 requiresnamespace,modified,temporal; L3 additionally requiresprovenanceandtemporal.validFrom.mif-convert emit-markdownschema-checks the JSON-LD input before projecting, unless--no-checkis given.roundtripreports whether md↔JSON-LD is lossless.hydrate-schemaresolveslatestby default and records the resolved version inschema/VENDOR.lock; offline,mif-validatefalls back to the last hydrated copy and warns.planner-checkwith no argument checks every recipe underskills/doc-set-planner/recipes/.
See also
Section titled “See also”The skills by purpose reference gives each skill its own
detailed write-up, grouped by the job it does; this catalog is the terse lookup
for genres, recipes, and scripts. The getting-started tutorial walks through
installing and validating a first document; the how-to gives the
validate/author/convert recipe; the explanation
covers why each document is both a human artifact and a machine unit. All three
are linked from this reference’s relationships[], which keeps rationale out of
the catalog per Diátaxis.