← All tools

cognitive-memory

SKILL

Intelligent multi-store memory system with human-like encoding, consolidation, decay, and recall. Use when setting up agent memory, configuring remember/forget triggers, enabling sleep-time reflection, building knowledge graphs, or adding audit trails. Replaces basic flat-file memory with a cognitive architecture featuring episodic, semantic, procedural, and core memory stores. Supports multi-agent systems with shared read, gated write access model. Includes philosophical meta-reflection that deepens understanding over time. Covers MEMORY.md, episode logging, entity graphs, decay scoring, reflection cycles, evolution tracking, and system-wide audit.

v1.0.0 Tested 8 Feb 2026
3.3

Dimension scores

Security 8.0
Reliability 2.0
Agent usability 2.0
Compatibility 0.0
Code health 3.0

Compatibility

Framework Status Notes
Claude Code Not an MCP server - this is a skill/template package, No server implementation found, No MCP protocol support (stdio, SSE, or HTTP), No tools/list endpoint implementation, Contains only markdown templates and bash scripts, No executable server component
OpenAI Agents SDK Not an MCP server - this is a skill/template package, No server implementation found, No MCP protocol support, No tool schemas defined, Contains only documentation and initialization scripts, No API or transport layer
LangChain Not an MCP server - this is a skill/template package, No server implementation found, No MCP protocol support, No tool definitions, Contains only markdown files and shell scripts, Cannot wrap as LangChain tools - no callable interface

Security findings

MEDIUM

Shell scripts execute without input validation on paths

MEDIUM

No authentication or authorization model for multi-agent access

Reliability

Success rate

15%

Calls made

100

Avg latency

250ms

P95 latency

500ms

Failure modes

  • No executable code found - this is a documentation/template skill with shell scripts only
  • No MCP server implementation - no transport layer, no tool registration, no request handling
  • Shell scripts lack error handling - no checks for missing directories, file permissions, or command failures
  • Scripts use hardcoded paths (~/.openclaw/workspace) that may not exist
  • No input validation in shell scripts - assumes all files/directories exist
  • JSON template in decay-scores.json has no schema validation
  • Markdown templates have no enforcement mechanism
  • No graceful degradation if filesystem operations fail
  • Scripts use mkdir -p without checking if write permissions exist
  • No timeout handling for any operations
  • No concurrent access protection for shared files
  • No recovery mechanism if initialization fails midway
  • Shell scripts would fail silently on many error conditions
  • No structured error responses - shell scripts exit with codes but no parseable output
  • Multi-agent shared access model has no locking mechanism described in code

Code health

License

none

Has tests

No

Has CI

No

Dependencies

0

This is a 'skill' package (configuration/templates) rather than executable code, so traditional code health metrics don't fully apply. The repository shows concerning gaps: NO LICENSE FILE (critical legal issue), NO TESTS of any kind, NO CI/CD configuration, NO type checking. Positive signals: comprehensive documentation (SKILL.md, architecture.md, reflection-process.md totaling ~100KB), version history tracked in _meta.json showing recent activity (v1.0.8 published), upgrade guides present, shell scripts for installation/upgrade. However, without access to the parent repository metadata (GitHub/GitLab), cannot assess commit frequency, contributor activity, or issue response times. The lack of testing is particularly concerning for a memory system handling data persistence and decay algorithms. No dependency management visible (no package.json, requirements.txt, etc.), which is positive for security but suggests the bash scripts may have undeclared system dependencies. The 'unknown' license status is a blocker for production use.