emasoft/amama-session-memory
skills/amama-session-memory/SKILL.md
Use when restoring user context, tracking decisions, or managing availability states. Trigger with session start or preference changes. Loaded by ai-maestro-assistant-manager-agent-main-agent
testing
Updated Apr 26, 2026
$ install --global
skillsauthnpx skillsauth add emasoft/ai-maestro-assistant-manager-agent amama-session-memoryInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 26, 2026, 3:56 AM16.5s6 files scanned
SKILL.md
- name:
- amama-session-memory
- description:
- Use when restoring user context, tracking decisions, or managing availability states. Trigger with session start or preference changes. Loaded by ai-maestro-assistant-manager-agent-main-agent
- version:
- 2.3.3
- compatibility:
- Requires AI Maestro installed.
- context:
- fork
- agent:
- amama-assistant-manager-main-agent
- user-invocable:
- false
AMAMA Session Memory Skill
Overview
Maintains user relationship continuity across sessions. Uses CozoDB (primary) with file-based fallback for preferences, decisions, pending items, and availability states.
Prerequisites
- AI Maestro with CozoDB at
~/.aimaestro/agents/<agent-id>/ - Subconscious memory indexing service running
- File access for
thoughts/shared/handoffs/amama/
Instructions
- Check CozoDB availability and initialize storage
- Load context from CozoDB or handoff files
- Run semantic search for implicit preferences
- Detect preference expressions and patterns
- Record decisions (approvals, rejections, deferrals)
- Track pending items awaiting user attention
- Monitor availability state changes
- Write handoffs before session end
- Trigger consolidation for stable preferences
Copy this checklist and track your progress:
- [ ] Check CozoDB / initialize storage
- [ ] Load context, detect preferences
- [ ] Record decisions, track pending items
- [ ] Write handoffs at session end
Output
| Output | When | |--------|------| | Session restored | Session start | | Preference recorded | Preference detected | | Decision logged | User decides | | Availability ack | State change | | Handoff record | Session end | | Pending reminder | Resume | | Sync status | CozoDB recovery |
Error Handling
| Error | Resolution | |-------|------------| | CozoDB unavailable | File-based fallback | | CozoDB write fail | Retry once, then fallback | | Fallback dir missing | Create directory | | File corrupted | Regenerate from CozoDB | | Semantic timeout | Structured data only | | Conflicting prefs | Most recent wins | | CozoDB/file diverge | CozoDB authoritative |
Examples
Session start: Loads preferences, pending items, decisions from CozoDB or fallback.
Preference detection: User says "be more brief" -- writes to CozoDB + fallback.
Input/Output: Store {"relation":"user_preferences","data":{"key":"response_length","value":"concise"}} => Output: Noted: keeping responses concise.
See references/examples.md for detailed examples:
- Example 1: Session Start with AI Maestro Memory
- Example 2: Session Start with Fallback
- Example 3: Detecting and Recording Preference
- Example 4: Recording Decision with Context
- Example 5: Availability State Change
- Example 6: Semantic Search for Implicit Preferences
Resources
- Related skills - amama-approval-workflows, amama-user-communication
- Memory - CozoDB with file fallback
Reference Documents
- memory-architecture.md - Storage tiers and schemas
- Tiered Storage Architecture
- Tier 1: AI Maestro CozoDB Agent Database
- Tier 2: Subconscious Memory Indexing
- Tier 3: Long-Term Memory Consolidation
- Tier 4: File-Based Handoff Documents (Fallback)
- Secondary: GitHub Issue Comments
- Data Types Stored
- memory-triggers.md - Trigger conditions and API
- Memory Retrieval Triggers
- Memory Update Triggers
- handoff-format.md (references/handoff-format.md) - Handoff templates
- examples.md - Operation examples
- Example 1: Session Start with AI Maestro Memory
- Example 2: Session Start with Fallback
- Example 3: Detecting and Recording Preference
- Example 4: Recording Decision with Context
- Example 5: Availability State Change
- Example 6: Semantic Search for Implicit Preferences
- record-keeping-formats.md (references/record-keeping-formats.md) - Record formats
Related Skills
emasoft/amama-user-communication
data-ai
VerifiedTrustedCommunity
Use when communicating with users for clarification, presenting options, requesting approval, or reporting completion. Trigger with user requests or communication needs. Loaded by ai-maestro-assistant-manager-agent-main-agent.
SKILL.mdUpdated Apr 26, 2026
emasoft/amama-status-reporting
development
VerifiedTrustedCommunity
Use when generating status reports. Queries AI Maestro APIs (sessions, agents, teams, tasks) for live data. Trigger with status report requests.
SKILL.mdUpdated Apr 26, 2026
emasoft/amama-role-routing
testing
VerifiedTrustedCommunity
Use when routing user requests to specialist agents. Trigger with work delegation needs. Loaded by ai-maestro-assistant-manager-agent-main-agent
SKILL.mdUpdated Apr 26, 2026
emasoft/amama-label-taxonomy
data-ai
VerifiedTrustedCommunity
Use when managing GitHub issue labels or priorities. Trigger with label or triage requests. Loaded by ai-maestro-assistant-manager-agent-main-agent.
SKILL.mdUpdated Apr 26, 2026