△
DMCallv1
Projects
AI Models
Docs
Inventory
Rules
Backup
AI Logs
Workspaces
Rules
/ AI Session Protocol / Edit
Edit Rule
Rule Info
Title *
Slug
Category
Workflow
Coding
Deployment
Database
Ai Usage
Backup
General
Sort Order
Public (accessible without API key)
Content (Markdown)
Markdown Source
# AI Session Protocol When an AI model starts a new session working on a project in this ecosystem: ## Before Starting Work 1. **Fetch project context:** ``` GET https://mssql.danmarcrm.com/dev1/dmcallv1/api/projects/{project-slug} ``` This gives all locations, SSH parameters, access methods, tech stack, and links. 2. **Read the rules** (you are reading one now): ``` GET https://mssql.danmarcrm.com/dev1/dmcallv1/api/rules ``` 3. **Check existing docs for the project:** ``` GET https://mssql.danmarcrm.com/dev1/dmcallv1/api/docs?project={project-slug} ``` ## During Work - Write discoveries (file paths, configs, patterns) to the inventory: ``` POST https://mssql.danmarcrm.com/dev1/dmcallv1/api/inventory X-Api-Key: {api-key} ``` ## After Work - If you created or discovered significant architectural information, update the relevant Doc entry. - Store the session log if it was a substantial conversation: ``` POST https://mssql.danmarcrm.com/dev1/dmcallv1/api/ailogs X-Api-Key: {api-key} ``` ## Code Locations Code may exist in multiple places: - **Server:** `mssql.danmarcrm.com` under `/var/www/mssql.danmarcrm.com/dev1/{project}/` - **Local (developer):** On the developer's machine (check project locations in DMCallv1) - Always check the `locations` array in the project record before assuming a file path.
Preview
Save Changes
Cancel