Skillget

cman

skill
Security Audit
Pass
Health Pass
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Community trust — 23 GitHub stars
Code Pass
  • Code scan — Scanned 11 files during light audit, no dangerous patterns found
Permissions Pass
  • Permissions — No dangerous permissions requested
Purpose
This is a Python-based CLI plugin that gives Claude agentic search capabilities over your past sessions, plans, and local memory files. It allows you to ask natural language questions about your coding history and summarizes the results.

Security Assessment
Overall Risk: Low. The automated code scan evaluated 11 files and found no dangerous patterns, hardcoded secrets, or requests for dangerous permissions. The tool requires no API keys and makes no external network requests. It operates entirely within your local environment, querying existing local data rather than sending data to external databases. While it does access sensitive data (your local session logs, plans, and development history), it is designed to read and summarize this information rather than exfiltrate it.

Quality Assessment
The project is actively maintained, with its most recent push occurring today. It is protected by the highly permissive and standard MIT license, making it safe for both personal and commercial use. While the repository is relatively small and targeted, it has garnered 23 GitHub stars, indicating a baseline level of community trust and usefulness. The documentation is clear, providing straightforward installation instructions and well-structured examples.

Verdict
Safe to use.
SUMMARY

Agentic Memory for Claude Code

README.md

cman — Agentic Memory for Claude Code

Your memory is already there. cman just searches it agentically.

No external database, no API keys, no extra storage. cman gives Claude search over your existing sessions, plans, and memory files — and lets it summarize what matters.

Installation

/plugin marketplace add laiso/cman
/plugin install cman@cman

How it works

Just ask Claude naturally:

> What did I do yesterday?
> What did I work on last week?
> Where was that auth-related work?

Claude automatically searches your sessions, plans, and memory files — then summarizes the results.

You can also use skills directly:

Skill Scope Purpose
/remember … All projects Search and summarize past sessions (e.g. /remember FOO BAR for keywords)
/cm-status Current project Quick overview of plans, sessions, and memory (all for everything)

The remember skill is the primary search command (/remember). With the plugin enabled, the palette may show the namespaced form (e.g. cman:remember). Workflow text lives in skills/cm-search/SKILL.md for now; the legacy cm-search skill name is deprecated and will be removed in a future release. cm-status may appear as cman:cm-status in the same way.

Example output

"What did I do yesterday?"

Show example 
## Work Recap — Past 24 hours

### Completed
- **web-app**: User dashboard chart component
- **web-app**: Fixed pagination bug on /users endpoint
- **web-app**: PR review (add rate limiter, update deps)
- **infra**: Cluster upgrade plan drafted
- **mobile-app**: Build config cleanup

### In Progress
- **web-app**: CSV export feature — `claude --resume a2ac93f1...`
- **infra**: Cost analysis — `claude --resume 06365b61...`

### Stats
- 50 sessions across 4 projects

"Where was that auth work?"

Show example 
## Search Results for "auth"

### Plans
- [Kubernetes cluster upgrade] — 3 sessions
  (Contains service account and auth configuration)

### Memory
- [web-app] project_oauth_migration.md
  OAuth2 migration — moving from session-based auth to JWT tokens

### Conversation matches (deep search)
- [web-app] Permission settings investigation — 5 days ago
  `claude --resume a82548ab...`

/cm-status

Show example 
Status

Active Plans
┌───┬──────────────────────────────────────────┬──────────┬─────────┐
│ # │ Plan                                     │ Sessions │ Project │
├───┼──────────────────────────────────────────┼──────────┼─────────┤
│ 1 │ feat: Add user dashboard with charts     │ 12       │ web-app │
│ 2 │ CSV export → Google Sheets integration   │ 16       │ web-app │
│ 3 │ Kubernetes cluster upgrade               │ 3        │ infra   │
│ 4 │ Frontend i18n support                    │ 11       │ web-app │
└───┴──────────────────────────────────────────┴──────────┴─────────┘

Recent Sessions
┌───┬─────────────────────────┬─────────┬────────────────┬──────────────────────────────────────────────────┬─────────────────────────────┐
│ # │ Title                   │ Project │ When           │ CLI Resume                                       │ Interactive                 │
├───┼─────────────────────────┼─────────┼────────────────┼──────────────────────────────────────────────────┼─────────────────────────────┤
│ 1 │ README editing          │ web-app │ 10 seconds ago │ cd ~/work/web-app && claude --resume a1b2c3d4... │ /resume a1b2c3d4...         │
│ 2 │ Status skill testing    │ web-app │ 1 minutes ago  │ cd ~/work/web-app && claude --resume e5f6g7h8... │ /resume e5f6g7h8...         │
│ 3 │ Search skill testing    │ web-app │ 3 minutes ago  │ cd ~/work/web-app && claude --resume i9j0k1l2... │ /resume i9j0k1l2...         │
└───┴─────────────────────────┴─────────┴────────────────┴──────────────────────────────────────────────────┴─────────────────────────────┘

Memory Overview
- web-app: 3 files
  - ~/.claude/projects/.../web-app/memory/MEMORY.md
  - ~/.claude/projects/.../web-app/memory/project_design.md
  - ~/.claude/projects/.../web-app/memory/feedback_conventions.md

Requirements

  • uv (Python package runner)

Architecture

Natural language          Skills (UX)              MCP Server            Scripts (data)
                    ┌─────────────────┐      ┌──────────────────┐  ┌────────────────┐
"What did I do      │ remember        │─────▶│ list_sessions    │──│ sessions.py    │
 yesterday?"     ──▶│ (auto-trigger)  │─────▶│ list_plans       │──│ plans.py       │
                    ├─────────────────┤─────▶│ list_memory      │──│ memory.py      │
/cm-status       ──▶│ cm-status       │─────▶│ search_sessions  │──│ grep.py        │
/remember …      ──▶│ remember        │─────▶│ (same pipeline)  │──│ (same)         │
                    └─────────────────┘      └──────────────────┘  └────────────────┘
                    (workflow text: skills/cm-search/SKILL.md; cm-search skill deprecated)

Skills call MCP tools served by server.py (stdio transport, launched via uv run --script). The MCP server reuses logic from the Python scripts, which read ~/.claude/projects/**/*.jsonl and memory files directly. Claude interprets, filters, and summarizes the output.

Development

# Local testing
claude --plugin-dir /path/to/cman

License

MIT

Reviews (0)

No results found