Stats
Actions
Tags
How this command is triggered — by the user, by Claude, or both
Slash command
/homunculus:add-sourceThe summary Claude sees in its command listing — used to decide when to auto-load this command
# Add Source Add a project path to the homunculus observation sources. ## Detect Current Project First, check if current working directory has a `.claude/` directory: ## Behavior **If user provides a path argument:** Use that path. **If no argument AND current directory has `.claude/`:** Offer to add the current directory. **If no argument AND no `.claude/` in current directory:** ## Add the Source ## Handle Result **ADDED:** **EXISTS:** **INVALID:** ## After Adding Suggest next steps:
Add a project path to the homunculus observation sources.
First, check if current working directory has a .claude/ directory:
test -d ".claude" && echo "HAS_CLAUDE" || echo "NO_CLAUDE"
If user provides a path argument: Use that path.
If no argument AND current directory has .claude/:
Offer to add the current directory.
If no argument AND no .claude/ in current directory:
No .claude/ directory found in current project.
Specify a path: /homunculus:add-source /path/to/project
"${CLAUDE_PLUGIN_ROOT}/scripts/add-source.sh" "<path>"
ADDED:
Added: /path/to/project
Observations from this project will be included in /homunculus:aggregate.
EXISTS:
Already configured: /path/to/project
INVALID:
Path does not exist: /path/to/project
Suggest next steps:
Use /homunculus:sources to see all configured sources.
Use /homunculus:aggregate to collect observations.
npx claudepluginhub cyotee/cyotee-claude-plugins --plugin homunculus/init-projectConfigures repo for Obsidian PKM by linking to vault project folder, optionally scaffolding index, and adding PKM directives to CLAUDE.md.
/discoverGenerates a structured project note by walking the project tree, reading manifests and docs, and invoking an LLM to produce a summary of stack, architecture, and conventions. Appends to existing notes without overwriting.
/add-sourceAdds Instagram handles or web aggregator URLs to sources.yaml config file. Profiles web sources with Firecrawl before saving; skips duplicates.