🌳 trigger-tree

See which docs your AI actually discovers —
and which project knowledge it never finds.

100% local · zero model tokens · no cloud, no analytics vendors

tt watch — live doc-discovery 

    
every flash is a real doc read rippling up the tree · ← older prompts · → newer prompts · this demo is the plugin

  


  

    
🔥 Heat & cold maps

      
Attention cools with a 30-day half-life while lifetime reads stay visible. Focused top 10 live; exact 7/30/90-day windows in insights. Captures rg/grep/find and expanded Bash reads.

    
📈 Proof, not vibes

      
/tt suggestions proposes fixes with evidence. /tt note marks the change; the trend shows whether it worked.

    
🔒 Local by design

      
Paths and prompt hashes stay local; prompt text is opt-in. Stdlib only, zero network calls. Delete .trigger-tree/ and it's gone.

    
🅰️ A grade your PO gets

      
Coverage, router gaps and hunting distilled into one documentation health score — trackable sprint over sprint.

    
⚡ Zero model tokens

      
Hooks run shell-side in milliseconds. Measuring costs nothing — not in tokens, not in latency, not in context.

    
🤝 Claude Code · Codex

      
Native lifecycle hooks feed one local dataset on both agents. Full CI covers macOS, Linux, Windows and Python 3.10–3.13; the dashboard opens in an iTerm2 split, tmux pane, or Windows Terminal.

  


  
Why this matters

  
In AI-assisted development, documentation is the steering wheel:
  it tells the assistant how your team builds software. But a rule that is never
  read protects nothing — an unread guardrail fails silently, and you only notice when the AI "ignores" a
  convention it simply never found. trigger-tree measures what actually gets read, flags what never does, and
  proves whether your fixes worked. Documentation becomes monitored infrastructure instead of a hopeful artifact.

  
And this metric exists nowhere else: agent observability platforms measure
  tokens and traces — none measure which of your docs the assistant actually read.
  Anthropic's own guidance says to keep CLAUDE.md lean; trigger-tree provides evidence for what to review,
  what to protect or reroute, and what to rescue.


  
Install

  

    

      
Claude Code

      
Slash commands, Claude hooks, and optional Claude statusline.

      
/plugin marketplace add Hedde/trigger_tree
/plugin install trigger-tree@trigger-tree
/tt setup
/tt doctor

    

    

      
OpenAI Codex

      
Codex lifecycle hooks and natural-language workflows. Codex's built-in /statusline is separate.

      
codex plugin marketplace add Hedde/trigger_tree
codex plugin add trigger-tree@trigger-tree
# start a new thread
# review and trust hooks with /hooks

    

  

  
Where it appears: this GitHub marketplace installs immediately under your configured
  marketplace/Installed plugins. It will appear under OpenAI Curated only after a separate OpenAI submission,
  review, approval, and publish step.


  
Use it

  
Claude Code uses the /tt commands below. In Codex, ask for the same outcome—for
  example “Show trigger-tree status” or “Open the trigger-tree live dashboard.”

  

    
CommandDoes

    
/tt statusSnapshot: reads, hot files, untouched paths

    
/tt watchThis dashboard, live, in your terminal

    
/tt insightsHeat/cold map analysis + HTML report

    
/tt suggestionsMax 5 evidence-backed router fixes

    
/tt noteAnnotate router changes on the timeline

    
/tt doctorProve this repo is wired and receiving events

    
/tt setupWire it into any project in one command