# cogsession

> cogsession — premanand8800-cogsession. Use this tool when you need to track and manage AI coding decisions, assumptions, and dead ends in real-time. Cogsession records and makes searchable the development process, alerting you when code validity changes, and integrates with git for version control. It's ideal for AI coding agents requiring transparent and auditable session memory, such as Claude Code on an MCP server.

Canonical page: https://skillsregistry.net/skills/premanand8800-cogsession  
JSON: https://api.skillsregistry.net/v1/skills/premanand8800-cogsession

## Description

Session memory for AI coding agents. Records decisions, dead ends and assumptions as they happen, makes them greppable, and tells you when what you wrote stops being true. MCP server for Claude Code.

## Trust

- **Trust score (0–1):** 0.70
- **Verification tier:** verified
- **Last scanned:** 2026-09-28

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** container
- **Runtime environment:** vm
- **License:** Apache-2.0
- **Updated:** 2026-09-28

## Source

- **Source listing:** [GitHub](https://github.com/premanand8800/cogsession)

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "premanand8800-cogsession"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/premanand8800-cogsession` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/premanand8800-cogsession/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt