Repo Glossary - Context Intake Metadata Review
generated: 2026-06-12
working packet:
research/_working/preliminary-repo-glossary-research.mdcanonical glossary: blocked until final compiled YAML approves the starter terms and output path.
Summary
This page renders the refreshed `$repo-glossary` working packet for the metadata terminology migration. The repo now uses context_intake as the canonical frontmatter field for user/context intake, with public labels Deep interview, Scoped intake, and Artifact-driven. The page asks whether the starter glossary should define these terms, preserve type as the broad workflow category, and include visual_tier as related skill metadata.
starter terms proposed
acronym proposed
source files referenced
canonical glossary files written
This is intentionally still a review page. Per the staged glossary workflow, research/glossary.md should be created only after final compiled YAML from this page returns approval_status: ready-for-agent-review with no unresolved negative feedback.
Table Of Contents
Scope Resolution
The active term family is repo-level skill metadata, not product-specific Skills Showcase research. The proposed approved artifact is therefore the parent glossary research/glossary.md. No scoped glossary is proposed for this run.
| Item | Decision | Reason |
|---|---|---|
| Product path | Parent glossary mode | context_intake, type, and visual_tier are shared skill metadata fields. |
| Working packet | research/_working/preliminary-repo-glossary-research.md | Non-canonical Stage 2 packet for review. |
| Canonical output | research/glossary.md | Created only after final compiled YAML approval. |
| Status values | confirmed only after approval | Starter glossary rows should not be marked confirmed before this page is approved. |
Starter Terms
| Term | Proposed definition | Category | Source |
|---|---|---|---|
Frontmatter | The YAML metadata block at the top of SKILL.md that declares machine-readable skill fields before the prose contract. | technical | docs/skill-anatomy.md |
Skill metadata | Structured fields parsed from SKILL.md frontmatter and used by docs, generators, catalogs, audits, and showcase data. | technical | docs/skill-anatomy.md, scripts/catalog/index.mjs |
type | The broad workflow category for a skill, preserved as one of planning, research, analysis, execution, review, shipping, ops, or router. | workflow | docs/skill-anatomy.md |
context_intake | The canonical skill frontmatter field that declares the expected user/context discovery cadence before substantive work; valid values are deep, scoped, and artifact_only. | workflow | docs/interview-convention.md, scripts/upgrade-alignment-page.mjs |
visual_tier | The skill frontmatter field that declares the expected alignment-page rendering tier: document, visual, or prototype. | workflow | docs/skill-anatomy.md, docs/alignment-page-convention.md |
| Deep interview | Public label for context_intake: deep, meaning broad upfront discovery, assumptions validation, and checkpoints before durable output. | workflow | docs/interview-convention.md |
| Scoped intake | Public label for context_intake: scoped, meaning focused clarification for bounded tasks or known artifacts before execution. | workflow | docs/interview-convention.md |
| Artifact-driven | Public label for context_intake: artifact_only, meaning the skill primarily works from supplied artifacts and visible context; it does not require a pre-work discovery interview, but may still ask blocker questions. | workflow | docs/interview-convention.md |
Acronym proposed: YAML - YAML Ain't Markup Language; the structured data syntax used for SKILL.md frontmatter and alignment-page approval payloads.
Retired Language
| Term or wording | Current decision | Follow-up impact |
|---|---|---|
interview_depth | Retired active frontmatter key. | Keep only migration notes and regression tests; use context_intake in active skills. |
full, light, none | Retired intake values mapped to deep, scoped, and artifact_only. | Use the new values in docs, generated guidance, and catalog data. |
| Old public intake wording | Avoid. Use product-facing labels. | Use Deep interview, Scoped intake, or Artifact-driven. |
Evidence Matrix
| Claim | Evidence | Inference | Confidence | Decision impact |
|---|---|---|---|---|
context_intake is canonical. | Docs define it; generator and catalog consumers read it; tests block the retired key. | The old field should not remain a long-term alias. | High | Glossary should define only the live field. |
type remains broad category metadata. | docs/skill-anatomy.md lists type categories separately from intake cadence. | Workflow category and interaction cadence are independent axes. | High | Glossary should prevent future misuse of type for intake behavior. |
visual_tier belongs in starter metadata terms. | Alignment convention, generator, catalog types, and parity audit now expose or validate it. | It is part of the same discoverable frontmatter vocabulary family. | High | Glossary should define it next to context_intake. |
| Canonical write should wait for final YAML. | The repo-glossary staged workflow requires review before canonical glossary writes. | Writing confirmed glossary rows now would skip approval. | High | This page asks for approval before creating research/glossary.md. |
Proposed Canonical Write
After approval, create research/glossary.md in the flat glossary format with Terms, Acronyms, and Recently Added tables. Each approved starter term should use Status confirmed, Source Skill $repo-glossary, and Approved In alignment/repo-glossary-skill-conventions.html.
| File | Timing | Mutation |
|---|---|---|
research/_working/preliminary-repo-glossary-research.md | Now | Refreshed with starter terms and evidence. |
alignment/repo-glossary-skill-conventions.html | Now | Updated to render the refreshed packet and approval gates. |
research/glossary.md | After final approval | Create parent glossary with confirmed starter rows and the YAML acronym. |
docs/history/archive/YYYY-MM-DD/HHMMSS/research/_working/preliminary-repo-glossary-research.md | After final approval | Archive the working packet and remove the active packet. |
Approval Questions
Q1 - Approve the eight starter terms and one acronym?
Q2 - Approve the parent glossary output path?
Q3 - Approve confirmed status for the starter glossary rows after creation?