create-sentence-markup
Grammarly-like sentence decomposition with NVIS confidence colors. Decomposes question text into annotated fragments showing which entities resolved, which are misspelled, which are fabricated, and which are unknown. Composes /extract-entities for grounding data and /interview for "Did you mean?" clarify prompts.
Informations de source
- Dépôt
- grahama1970/agent-stack-public
- Dernière activité de la source
- 24 septembre 2026 à 15:51
- Langue détectée de SKILL.md
- anglais
- Étoiles
- 0
- Forks
- 0
Options d'installation
Le prompt qui vérifie d'abord la source est sélectionné par défaut. Vous pouvez passer à une commande directe ou télécharger une copie locale.
Vérifiez les fichiers source
Lisez SKILL.md et les fichiers associés affichés par SkillsMP avant de décider de l'installer.
Explorateur de fichiers
4 fichiersAffichage de SKILL.md
SKILL.md
Instructions source · Aperçu en lecture seule- name
- create-sentence-markup
- description
- Grammarly-like sentence decomposition with NVIS confidence colors. Decomposes question text into annotated fragments showing which entities resolved, which are misspelled, which are fabricated, and which are unknown. Composes /extract-entities for grounding data and /interview for "Did you mean?" clarify prompts.
- allowed-tools
- ["Bash","Read"]
- triggers
- ["markup sentence","annotate sentence","sentence markup","create sentence markup","grammarly","highlight entities"]
- metadata
- {"short-description":"NVIS-colored sentence annotation from entity grounding","author":"Claude","version":"0.1.0"}
- provides
- ["sentence-markup"]
- composes
- ["extract-entities","interview","agentic-evals"]
- disciplines
- ["extraction","ui-design-engineering"]
# /create-sentence-markup
Grammarly-like sentence decomposition with NVIS confidence colors (MIL-STD-3009).
## Usage
```bash
# Annotate a question — returns JSON annotations
./run.sh annotate "How does SPARTA control X23-MUSTARD mitigate spoofing?"
# Annotate with rendered markdown output
./run.sh annotate "How does the SPRTA framework work?" --format markdown
# Annotate with HTML output (NVIS colors)
./run.sh annotate "What countermesures protect firmware?" --format html
# Pipe from /extract-entities (skip redundant extraction)
./run.sh annotate --entities-json entities.json "How does X23-MUSTARD work?"
```
## Output (JSON default)
```json
{
"text": "How does SPARTA control X23-MUSTARD mitigate spoofing?",
"annotations": [
{
"term": "X23-MUSTARD",
"level": "RED",
"label": "fabricated ID — not in corpus",
"action": "reject",
"closest_match": "CM0028"
},
{
"term": "SPARTA",
"level": "GREEN",
"label": "confirmed framework",
"action": null
}
],
"summary": {
"total_annotations": 2,
"red": 1,
"amber": 0,
"yellow": 0,
"green": 1,
"needs_clarify": 0,
"needs_reject": 1
}
}
```
## NVIS Color System (MIL-STD-3009)
| Level | Color | RGB | Meaning | Action |
|-------|-------|-----|---------|--------|
| GREEN | Green | (0,255,136) | Exact match — confirmed in corpus | None |
| AMBER | Amber | (255,170,0) | Fuzzy match or misspelling | /memory clarify via /interview |
| RED | Red | (255,68,68) | Fabricated ID — not in corpus | Reject |
| YELLOW | Yellow | (255,230,0) | Term not found anywhere | Investigate |
## Composition
- **Input**: Question text (string)
- **Depends on**: `/extract-entities` → `get_annotations()` for grounding data
- **Triggers**: `/interview` for AMBER "Did you mean?" clarify prompts
- **Consumed by**: `/create-evidence-case` (report grounding), `/ask` (inline annotations),
`/lean4-prove` (proof obligations from RED/YELLOW annotations)
## Output Formats
| Format | Flag | Use case |
|--------|------|----------|
| JSON | `--format json` (default) | Machine-readable, piping to other skills |
| Markdown | `--format markdown` | Agent/human readable in terminal |
| HTML | `--format html` | Rich rendering with NVIS colors |
Voir sur GitHub