Our review
Extracts Federated Taxonomy tags from text for memory storage and graph traversal, using LLM extraction and deterministic validation.
Strengths
- Provides structured tags (bridge, collection) for cross-collection queries.
- Uses deterministic validation to filter to known vocabulary.
- Includes confidence and worth_remembering fields for filtering.
- Supports multiple collection types (lore, operational, sparta).
Limitations
- Requires a predefined taxonomy vocabulary which may not cover all concepts.
- The script (run.sh) needs to be available and configured.
- Only works with text input, not images or audio.
When you need to extract standardized tags from textual content to store in a memory graph and enable graph-based retrieval.
When you need raw keyword extraction without taxonomy constraints, or when the content is not text-based.
Security analysis
SafeThe skill uses Bash and Read tools to run a tag extraction script with deterministic validation against known vocabularies. There are no destructive commands, exfiltration attempts, or obfuscation. The functionality is benign and well-defined.
No concerns found
Examples
Extract Federated Taxonomy tags from this text: 'Error handling in the authentication module' with collection operational.Extract bridge tags from this text: 'The system is resilient against failures, but has some fragile legacy components' and only return bridge tags.Classify this narrative text with taxonomy tags: 'The hero's journey through the corrupted data maze' for lore collection.name: taxonomy description: > Extract Federated Taxonomy tags from text. LLM extracts candidates, deterministic validation filters to known vocabulary. Returns bridge tags for multi-hop graph traversal. allowed-tools: ["Bash", "Read"] triggers:
- taxonomy
- tag this
- classify
- what tags
- bridge tags metadata: short-description: Extract taxonomy tags for graph traversal
Taxonomy
Extract Federated Taxonomy tags from text for memory storage and multi-hop graph traversal.
Quick Start
# Extract tags from text
./run.sh --text "Error handling in the authentication module" --collection operational
# Extract from file
./run.sh --file document.txt --collection lore
# Just get bridge tags (for graph traversal)
./run.sh --text "..." --bridges-only
Output
{
"bridge_tags": ["Resilience", "Loyalty"],
"collection_tags": {"function": "Fix", "domain": "Middleware"},
"confidence": 0.8,
"worth_remembering": true
}
Bridge Tags (Tier 0)
Shared across all collections - enable cross-collection queries:
| Tag | Indicates | |-----|-----------| | Precision | Methodical, optimized, algorithmic | | Resilience | Fault tolerance, error handling, robustness | | Fragility | Technical debt, brittleness, single point of failure | | Corruption | Bugs, data corruption, silent failures | | Loyalty | Security, compliance, auth, encryption | | Stealth | Hidden, evasion, infiltration |
Collection Types
- lore - Narrative/story content (HLT vocabulary)
- operational - Code/technical lessons (Operational vocabulary)
- sparta - Security content (ATT&CK/D3FEND vocabulary)
Use Cases
- Before storing to memory: Get tags to enable graph traversal
- Filtering what to remember: Check
worth_rememberingfield - Cross-collection queries: Use bridge_tags for multi-hop search
Composing with /learn
# /learn can call /taxonomy for tag extraction
tags=$(./run.sh --text "$content" --collection operational --json)
bridge_tags=$(echo "$tags" | jq -r '.bridge_tags | join(",")')
Prompt Engineering
Data & AI
Prompt engineering best practices and templates to maximize AI outputs.
Data Visualization
Data & AI
Generates data visualizations and charts tailored to your data.
RAG Architecture Setup
Data & AI
Setup guide for RAG (Retrieval-Augmented Generation) architectures.