Skip to main content

document-spf-feature

Document an SPF feature record. Use when the user explicitly requests capability scope, relationships, status, constraints, or evidence.

Jump to install

Source facts

Repository
videojs/v10
Last source activity
August 25, 2026 at 20:08
Detected SKILL.md language
English
Stars
947
Forks
91

Install options

The review-first prompt is selected by default. You can switch to a direct command or download a local copy.

Review the source files

Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.

File Explorer
2 files

Showing SKILL.md

SKILL.md
Source instructions · Read-only preview
name
document-spf-feature
description
Document an SPF feature record. Use when the user explicitly requests capability scope, relationships, status, constraints, or evidence.
# Document an SPF feature Read `references/workflow.md` completely before acting; it contains the detailed registry, decomposition, evidence, cascade, and validation workflow. Confirm that the user explicitly requested creation or revision of the entry. Do not create adjacent feature, use-case, design, or decision records discovered during the workflow; surface them as candidates instead. Registry entries guide planning but do not override code. Read `internal/design/spf/features/clusters.md`, a strong neighboring entry, relevant implementation/tests, and linked records. 1. Classify the engine capability and distinguish it from delivery use cases or implementation conventions. 2. Separate current behavior from proposed direction and decisions still needed. 3. For shipped work, retain decisions, consequences, and current source pointers. For future work, retain scope, boundaries, and evidence required before implementation. 4. Update directly affected entries only when their facts changed; verify links and relationship symmetry. Cite repository paths for implemented behavior. Remove phase tables, speculative file inventories, and progress logs once code lands. Do not mark work implemented without code and verification evidence, and do not implement the feature unless requested. ## Example Input: “Document the current depth of discontinuity handling.” Output: A compact evidence-backed entry separating implemented behavior, remaining decisions, constraints, and source pointers.
View on GitHub