Skip to content

config.mdx

Latest commit

 

History

History
51 lines (45 loc) · 2.23 KB

config.mdx

File metadata and controls

51 lines (45 loc) · 2.23 KB
Raw
title Configuration
description Where settings, prompts, and connections live on disk.

Settings are stored at ~/.coder/config.json. Created with defaults on first launch; editable directly to add providers or models without opening the settings UI.

{
  "api_keys": {
    "fireworks": "fw-...",
    "xai": ""
  },
  "providers": {
    "fireworks": {
      "default": "accounts/fireworks/models/deepseek-v4-flash",
      "models": [
        { "id": "accounts/fireworks/models/deepseek-v4-flash", "label": "DeepSeek-V4-Flash ($0.45)", "contextWindow": 164000 }
      ]
    }
  }
}

Where things live under ~/.coder/

Path Purpose
config.json API keys, model lists, editor preferences, keybindings
profiles.json Index of known profiles for multi-account switching, see Accounts & Profiles
harness.md Guardrails always appended to the system prompt, see AI Agent
prompts/*.md Custom prompts, see AI Agent
skills/<name>/SKILL.md Progressive-disclosure skills, see Skills
connections/databases.json Global database connections, see Databases
connections/sftp.json Global SFTP connections, see SFTP
themes/<id>/ui.css, editor.json Custom themes, see Themes
recorder.db Full request/response recordings, see Reports
global-board.json Master Board cards not tied to any project, see Master Board
projects/{id}/token-log/{date}.jsonl Per-round token logs, see Token Usage Dashboard
projects/{id}/debugger/{timestamp}.jsonl Recorded debug sessions, see Debugger
docs/ This documentation, mirrored from the public coder-docs repo

Per-project files

Path Purpose
{project-root}/board.json Kanban board, see Kanban Board
{project-root}/connections/databases.json Project-scoped database connections
{project-root}/connections/sftp.json Project-scoped SFTP connections
{project-root}/.coder/memory/*.md AI project memory, see AI Agent
{project-root}/.coder/orchestration/runs/<runId>/ Orchestrator run data, see Orchestrator