skill-recipes
Composable patterns for advanced skill usage including skill chaining, telemetry instrumentation, visual cataloging, and skill versioning. Use when building complex multi-skill workflows, adding observability to skills, visualizing skill relationships, or managing skill evolution.
Installation
Follow the repository instructions for this skill and your agent. Review the files and dependencies before adding it to your workflow.
View source ↗GuidesSkill instructions
Skill Recipes
Composable patterns for advanced skill usage in SEA-Forge™. Recipes combine multiple skills and patterns into reusable workflows.
For reference: Skill Creator | PM-Agent Orchestration
Recipe Catalog
| Recipe | Purpose | Ingredients |
|---|---|---|
| Skill Chaining | Connect skills in pipelines | skill-creator + pm-agent |
| Skill Telemetry | Observe skill execution | observability-debugging + semantic-anchoring |
| Visual Catalog | Map skill relationships | cognitive-artifacts-builder + semantic-anchoring |
| Skill Versioning | Track skill evolution | skill-creator + semantic-anchoring |
Recipe 1: Skill Chaining
Purpose: Connect skills in pipelines where one skill's output triggers another.
Pattern: Sequential Chain
# chain-definition.yaml
chain:
name: "spec-to-artifact"
description: "Generate cognitive artifact from specification"
steps:
- skill: "doc-coauthoring"
trigger: "user requests spec document"
output: "spec.md"
- skill: "semantic-anchoring"
trigger: "spec.md created"
input: "spec.md"
output: "anchored-spec.md"
- skill: "cognitive-artifacts-builder"
trigger: "anchored-spec.md created"
input: "anchored-spec.md"
output: "project-checklist.cadsl"
Pattern: Conditional Chain
chain:
name: "document-pipeline"
steps:
- skill: "doc-coauthoring"
output: "document.md"
- condition: "document.type == 'technical'"
then:
skill: "docx"
template: "technical-report"
else:
skill: "cognitive-artifacts-builder"
artifactType: "Notebook"
Pattern: Fan-Out/Fan-In
chain:
name: "multi-skill-analysis"
steps:
- skill: "research-agent"
output: "context.json"
- parallel: # Fan-out
- skill: "semantic-anchoring"
input: "context.json"
output: "concepts.yaml"
- skill: "knowledge-graph-query"
input: "context.json"
output: "relationships.json"
- skill: "cognitive-artifacts-builder" # Fan-in
inputs: ["concepts.yaml", "relationships.json"]
artifactType: "MindMap"
Implementation
Add to your skill's SKILL.md:
## Chain Triggers
This skill can be triggered by:
- `spec.md` file creation (from doc-coauthoring)
- `case.created` event (from case-management)
This skill outputs:
- `artifact.cadsl` → triggers cognitive-artifacts-builder
- `concepts.yaml` → triggers semantic-anchoring
Recipe 2: Skill Telemetry
Purpose: Instrument skills with OpenTelemetry for observability.
Pattern: Span per Skill Invocation
# skill_instrumentation.py
from opentelemetry import trace
from opentelemetry.trace import SpanKind
tracer = trace.get_tracer("sea-forge.skills")
def instrument_skill(skill_name: str):
"""Decorator to instrument skill invocations."""
def decorator(func):
async def wrapper(*args, **kwargs):
with tracer.start_as_current_span(
f"skill.{skill_name}",
kind=SpanKind.INTERNAL,
attributes={
"skill.name": skill_name,
"skill.version": "1.0.0",
"sea.component": "skill-engine",
}
) as span:
try:
result = await func(*args, **kwargs)
span.set_attribute("skill.success", True)
span.set_attribute("skill.output_type", type(result).__name__)
return result
except Exception as e:
span.set_attribute("skill.success", False)
span.set_attribute("skill.error", str(e))
span.record_exception(e)
raise
return wrapper
return decorator
# Usage
@instrument_skill("cognitive-artifacts-builder")
async def generate_artifact(request):
# Skill logic here
pass
Pattern: Metrics per Skill
from opentelemetry import metrics
meter = metrics.get_meter("sea-forge.skills")
# Counters
skill_invocations = meter.create_counter(
"skill.invocations",
description="Number of skill invocations",
unit="1"
)
# Histograms
skill_duration = meter.create_histogram(
"skill.duration",
description="Skill execution duration",
unit="ms"
)
# Usage
skill_invocations.add(1, {"skill.name": "cognitive-artifacts-builder"})
skill_duration.record(duration_ms, {"skill.name": "cognitive-artifacts-builder"})
Semantic Envelope Integration
# skill-telemetry-envelope.yaml
envelope:
traceId: "abc123"
spanId: "def456"
semantic:
skill:
name: "cognitive-artifacts-builder"
version: "1.0.0"
chain: "spec-to-artifact"
step: 3
artifact:
type: "Checklist"
conceptIds:
- "sea:Deployment"
- "sea:QualityGate"
case:
id: "case-2026-042"
stage: "execution"
Dashboard Query (OpenObserve)
-- Skill success rate
SELECT
json_extract(attributes, '$.skill.name') as skill,
COUNT(*) as total,
SUM(CASE WHEN json_extract(attributes, '$.skill.success') = true THEN 1 ELSE 0 END) as success,
ROUND(100.0 * SUM(CASE WHEN json_extract(attributes, '$.skill.success') = true THEN 1 ELSE 0 END) / COUNT(*), 2) as success_rate
FROM traces
WHERE json_extract(attributes, '$.sea.component') = 'skill-engine'
GROUP BY skill
ORDER BY total DESC
Recipe 3: Visual Skill Catalog
Purpose: Generate visual maps of skill relationships using Mermaid diagrams.
Pattern: Skill Dependency Graph
# catalog-definition.yaml
artifactType: SkillCatalog
metadata:
title: "SEA-Forge™ Skill Catalog"
generatedBy: "skill-recipes"
visualization:
type: "mermaid"
diagram: "flowchart"
Generated Mermaid:
flowchart TB
subgraph Core["Core Development"]
SC[skill-creator]
DC[doc-coauthoring]
DX[docx]
XL[xlsx]
PF[pdf]
end
subgraph Cognitive["Cognitive Architecture"]
CA[cognitive-artifacts-builder]
SA[semantic-anchoring]
KG[knowledge-graph-query]
end
subgraph Orchestration["Orchestration"]
CM[case-management]
PM[pm-agent-orchestration]
OD[observability-debugging]
end
subgraph Integration["Integration"]
MB[mcp-builder]
FE[fetch]
CV[convert]
end
SC --> CA
SC --> PM
SA --> CA
SA --> KG
PM --> CM
CA --> DX
OD --> PM
Pattern: Skill Chain Visualization
sequenceDiagram
participant U as User
participant DC as doc-coauthoring
participant SA as semantic-anchoring
participant CA as cognitive-artifacts-builder
participant DX as docx
U->>DC: Request spec document
DC->>DC: Gather context
DC->>SA: Validate concepts
SA->>SA: Check Knowledge Graph
SA-->>DC: Validated refs
DC->>CA: Generate checklist
CA->>CA: Build CADSL
CA->>DX: Export to Word
DX-->>U: Deliver document
Generate Catalog
Generate a skill catalog diagram:
just skill-catalog # Mermaid format (default)
just skill-catalog list # List format
just skill-list # Quick skill listing
Recipe 4: Skill Versioning
Purpose: Track skill evolution with semantic versioning and lineage.
Pattern: Version Metadata
# SKILL.md frontmatter
---
name: cognitive-artifacts-builder
version: 2.0.0
replaces: cognitive-artifacts-builder@1.5.0
changes:
- type: "feature"
description: "Added 5 new artifact types (Notebook, Tracker, Outliner, Dashboard, Scorecard)"
- type: "breaking"
description: "web-artifacts-builder merged into this skill"
semanticRefs:
- conceptId: "sea:SkillDefinition"
- conceptId: "sea:CognitiveArtifact"
---
Pattern: Lineage Tracking
lineage:
skill: "cognitive-artifacts-builder"
versions:
- version: "1.0.0"
date: "2025-06-15"
status: "deprecated"
- version: "1.5.0"
date: "2025-11-20"
status: "deprecated"
changes:
- "Added Timeline and Kanban artifacts"
- version: "2.0.0"
date: "2026-01-06"
status: "current"
replaces: "1.5.0"
absorbs: ["web-artifacts-builder@1.0.0"]
changes:
- "Added Notebook, Tracker, Outliner, Dashboard, Scorecard"
- "Merged web-artifacts-builder functionality"
- "Added CADSL to React mapping"
Pattern: Migration Guide
## Migrating from v1.5 to v2.0
### Breaking Changes
1. **web-artifacts-builder removed**
- Use `cognitive-artifacts-builder` instead
- Scripts are identical: `scripts/init-artifact.sh`
### New Features
1. **5 new artifact types**
- Notebook, Tracker, Outliner, Dashboard, Scorecard
- See "Artifact Type Catalog" section
### Deprecated
1. **None** - All v1.5 features carry forward
Check Skill Version
just skill-version-check cognitive-artifacts-builder
# Output:
# 🔍 Checking skill: cognitive-artifacts-builder
# Version: 2.0.0
# Replaces: 1.5.0
# ⚠️ Contains breaking changes
# - web-artifacts-builder merged into this skill
Combining Recipes
Recipes can be combined for complex workflows:
# combined-recipe.yaml
name: "instrumented-skill-chain"
recipes:
- skill-chaining:
chain: "spec-to-artifact"
- skill-telemetry:
spans: true
metrics: true
envelope: true
- visual-catalog:
include-chain: true
format: mermaid
References
Examples
Read the usage examples in the source repository.
View source ↗Source history
- Indexed on
- 2026-01-20
- Index updated
- 2026-01-26
- Branch
- main
Index and sync dates describe this directory. View the source repository for code history.
GitHub ↗chat Comments (0)
Sign in to join the discussion and leave a comment.
No comments yet. Be the first to share your thoughts!