JCMRS Marketplace for Claude Code Plugins
A curated marketplace of Claude Code plugins.
Available Plugins
Profile Creator
Status: 🧪 Testing Phase
Plugin extending the Axivo Claude Collaboration Platform, providing functionality to create Domain Profiles - rooted in behavioral programming, SRE, Domain Knowledge Graphs Ontology (langroid, crewai, langraph, semantic kernel et alii) - and comprehensive guided templates and validation.
Features:
- 6-phase guided creation methodology
- Extensive inline guidance for non-technical users
- Templates for singular and composite (HMAS) profiles
- Reference examples across multiple domains
- Complete validation and testing framework
Location: ./plugins/profile-creator
Running Log
Status: 🧪 Testing Phase (Phase 1 Complete, Phase 2 Prototype Testing In Progress)
Persistent schema-driven running log that captures ideas, consultations, and Claude's reasoning patterns for cross-session learning and process memory. Addresses context window dependency by maintaining structured decision trails, assumptions, and learnings across sessions.
Features:
- Automatic signal detection (5 regex patterns for uncertainty, assumptions, decisions, critical issues)
- Profile-driven automation (DEVELOPER, RESEARCHER, ENGINEER thresholds)
- 3-cadence monitoring (session start, mid-toolchain, session end)
- 3-layer noise filtering (confidence threshold, entry cap, deduplication)
- Auto-generated sections (High-Priority Ideas, Open Risks, Linked Process Insights)
- Variant support (research-running-log, architecture-running-log)
Commands:
/activate running-log — Initialize skill
/log [type] [content] — Manual entry creation
/review [filter] — Query and display log
Location: ./plugins/running-log
Procedural Memory System (PMS)
Status: 🧪 Testing Phase (v1.0.0 - Automated Tests Pass, Manual Validation Required)
"Shame on you." - Because Claude learns from your corrections.
Long-term learning system that transforms conversational history into actionable procedural knowledge through a three-tier memory architecture. Captures both rigid procedures (technical patterns, coding practices) and flexible processes (workflows, decision patterns), extracting recurring behaviors and synthesizing context-aware rules that persist across sessions.
Features:
- Three-tier memory architecture: Episodic → Semantic → Procedural transformation pipeline
- Dual encoding strategies: Context-first encoding + JSONL fallback for comprehensive session capture
- Intelligent pattern detection: Frequency-based analysis with configurable thresholds (emerging, strong, critical)
- Automatic rule synthesis: Generates markdown rules from strong patterns, injected into
.claude/rules/pms/
- Privacy-first design: Automatic redaction of API keys, passwords, tokens, and custom sensitive patterns
- Hook integration: Auto-encoding at PreCompact, SessionEnd, Stop, and SessionStart events
- Error recovery: Corruption detection, backup system, and rebuild from episodic records
- Comprehensive testing: 80 tests with 100% critical path coverage
Commands:
/pms:encode — Manual episodic encoding trigger
/pms:extract — Semantic pattern extraction from sessions
/pms:synthesize — Procedural rule generation from patterns
/pms:reflect — Complete workflow (encode → extract → synthesize)
/pms:status — Display memory system state
/pms:validate — Validate JSON schema integrity
/pms:reset — Clear semantic/procedural memory
/pms:rebuild — Regenerate semantic knowledge from episodic records
Documentation:
ARCHITECTURE.md - Technical deep dive into component design
SKILL.md - Plugin-dev integration guide
examples/USAGE_EXAMPLES.md - 7 workflow scenarios + troubleshooting
VALIDATION_GUIDE.md - End-to-end testing instructions
CHANGELOG.md - v1.0.0 release notes
Location: ./plugins/claude-pms
Documentation Reader
Status: 🧪 Testing Phase
Documentation Reader for Axivo Claude Collaboration Platform, providing access through components, protocols, and competencies. Provides systematic access to Claude Collaboration Platform documentation, enabling both users and Claude Code to query and reference official documentation organized by platform architecture.
Features:
- Architecture-Aware Documentation Access: Understands platform structure (Components, Protocols, Competencies)
- Dual Use Cases: Supports both user-initiated queries and agent self-reference
- Progressive Disclosure: Lean core skill with focused resource files
- On-Demand Fetching: Retrieves documentation from GitHub as needed
Commands:
/jcmrs:docs [search query] — Search official platform documentation