extend-protocol

name	extend-protocol
description	Extend the Scoutica Protocol with new schemas, entity types, rule templates, or CLI commands. Use when asked to add new features, modify the card format, add support for new entity types (agents, teams, robots), or contribute to the protocol specification.

extend-protocol

Add new capabilities to the Scoutica Protocol — new schemas, entity types, rules, or CLI features.

When to Use

• User wants to add a new field to the card schema
• User wants to support a new entity type (AI agent, team, robot)
• User wants to add a new CLI command
• User wants to create a new evaluation rule template
• User wants to modify the protocol specification

Extension Points

1. Add a New Card Field

The candidate card has no standalone JSON Schema file — profile.json is a flat object written by the CLI and checked by tools/validate_card.py. To add a field (e.g. a top-level ai_models array of models the candidate has trained or deployed), extend the flat profile shape:

 {
   "schema_version": "0.1.0",
   "skills": ["Python", "Go", "React"],
   "tools_and_platforms": ["Kubernetes", "AWS"],
+  "ai_models": ["Llama 3", "Stable Diffusion"]
 }

Checklist:

• Add the field to the writer in tools/scoutica (cmd_init() / the profile builder)
• Add validation for the field in tools/validate_card.py
• Update SCAN_PROMPT.md so AI generation includes the new field
• Update sample card in protocol/examples/sample_card/
• Update SKILL.md agent instructions
• Validate with scoutica validate <card-folder>

2. Add a New Entity Type

The protocol supports multiple entity types beyond humans:

Type	Card For	Example
human	Individual professional	Software engineer
ai_agent	Autonomous AI agent	Coding assistant, trading bot
service	API or SaaS product	Translation service, CDN
robot	Physical autonomous system	Warehouse robot, drone
team	Group of entities	Engineering squad
organization	Company or department	DevOps team at Company X

To add a new entity type:

1. Create a schema variant: schemas/<type>_profile.schema.json
2. Create a card template: protocol/platform/02_skill_patterns/template_<type>_skill.md
3. Update the base schema's entity_type enum
4. Add CLI support: scoutica init --type <type>

3. Add a New Rule Template

Location: protocol/templates/rules/

Create a new .md file that follows this pattern:

# Rule: <rule-name>

## Purpose
What this rule evaluates.

## Input
- `profile.json` fields used
- `rules.yaml` fields used

## Algorithm
1. Step-by-step evaluation logic
2. Clear decision criteria

## Output
- PASS / FAIL / SOFT_REJECT
- Reason string
- Confidence score (0-100)

Checklist:

• Create the rule file in protocol/templates/rules/
• Reference it in SKILL.md evaluation section
• Add to cmd_scan() so generated cards include it
• Update protocol/examples/sample_card/rules/

4. Add a New CLI Command

Source: tools/scoutica (bash) and tools/scoutica.ps1 (PowerShell)

# In tools/scoutica — add the function:
cmd_newcommand() {
    # Implementation here
}

# In the main router:
case "$command" in
    init)       cmd_init "$@" ;;
    scan)       cmd_scan "$@" ;;
    newcommand) cmd_newcommand "$@" ;;  # ← add here
    ...
esac

# In cmd_help():
echo -e "  ${CYAN}newcommand${NC}       Description of what it does"

Checklist:

• Implement in tools/scoutica (bash)
• Implement in tools/scoutica.ps1 (PowerShell)
• Add to cmd_help() in both CLIs
• Add to README.md commands section
• Update SKILL.md agent instructions
• Update install.sh / install.ps1 if new files needed

5. Add a New AI Provider for Scan

Source: tools/scoutica → cmd_scan() function

# In the auto-detection order:
if command -v gemini &>/dev/null; then
    provider="gemini"
elif command -v claude &>/dev/null; then
    provider="claude"
elif command -v newcli &>/dev/null; then   # ← add here
    provider="newcli"
fi

# In the provider case statement:
case "$provider" in
    newcli)
        response=$(newcli --prompt "$(cat "$tmp_prompt")" 2>/dev/null)
        ;;
esac

Development Workflow

1. Fork the repo
2. Create a feature branch
3. Make changes following the checklists above
4. Run scoutica validate on the sample card
5. Test with scoutica scan using your changes
6. Submit a PR with a clear description

File Map

What	Where
JSON Schemas	schemas/
CLI (bash)	tools/scoutica
CLI (PowerShell)	tools/scoutica.ps1
Card templates	protocol/templates/
Rule templates	protocol/templates/rules/
AI scan prompt	tools/SCAN_PROMPT.md
Sample card	protocol/examples/sample_card/
Protocol specs	protocol/platform/
Flow diagrams	.specs/protocol_flows.md
Agent instructions	SKILL.md