DocsIntegrationsFeed Pipeline

Feed Pipeline

Feed files and directories into HelixMind's spiral memory for the AI to learn from your codebase.

Feed Pipeline

The feed pipeline imports files in bulk into the spiral memory.

Usage

helixmind feed README.md src/architecture.md   # Specific files
helixmind feed src/                             # Directory (recursive)
/feed                                           # Interactive in chat

Feed Modes

ModeFlagDescription
Quick (default).Fast overview, key patterns
Deep--deepThorough analysis of every file
Watch--watchContinuous monitoring on file changes

What Gets Extracted

File TypeExtraction
.ts, .jsFunctions, classes, exports, imports
.jsonConfig structure, keys
.mdSections, concepts
Config filesFramework settings

Automatically ignored: node_modules/, dist/, .git/, binaries, lock files.

Memory Impact

Fed content starts at L2 Active. Over time: frequently used knowledge promotes to L1, unused decays to L3โ€“L5. Web-fetched knowledge enters separately at L6.

Tips

  1. Feed early: helixmind feed README.md right after helixmind init
  2. Deep-feed core code: --deep for your most important directories
  3. Don't over-feed: Quality over quantity
  4. Re-feed after major changes: After refactoring, re-feed the changed directories
HelixMind โ€” AI Coding CLI with Spiral Memory