Six open agent terminals look productive—even when four are waiting on a handoff and two are trapped in a rework loop. Dendro reads your existing local session logs afterward to show where time, handoffs, and budget actually went.
When you run coding agents across multiple terminals, worktrees, or team roles, you get megabytes of raw logs and very little clarity on what actually moved the work forward.
Dendro is built for engineering leaders, platform architects, and developers who want honest operational feedback rather than AI hype.
Instead of counting raw turns or lines of code, Dendro reconstructs the true timeline of every seat—separating active model and tool work from human wait times, peer handoff delays, and stalled queues.
When an agent's context window fills up and compacts, or when a reviewer rejects a change three times in a row, costs spike. Dendro highlights exactly where context drops and rework loops happen.
Dendro never invents "developer hours saved" or subjective code-quality scores. Every chart—in the terminal cockpit, the offline browser explorer, or plain-English queries—is backed by local log facts.
Dendro works standalone on any machine with local agent logs. When combined with the rest of the Move the Needle open-source stack, it closes the learning loop.
agent-otel-bridge is the live dashboard showing what your agents are doing right now in Grafana or Datadog. Dendro is the flight recorder you read afterward to understand cost, rework, and handoff flow.
Run parallel coding tasks inside isolated sandboxes with Meshloop, then use Dendro to inspect how long each task took and how many repair turns were needed.
Define clear engineering roles and acceptance gates with AgentFlow, and use Dendro to measure whether role handoffs are smooth or bouncing back in rework.
Built by Samuel Mota · Move the Needle
Open source · MIT · v0.7.0
See where time, handoffs, rework, and budget actually go across your AI coding sessions—using verifiable local evidence instead of vanity productivity claims.
Six open terminals look busy even when four agents are stalled on a handoff. Dendro turns raw session history into honest evidence of where time, budget, and rework actually go.
Reads local session transcripts across six AI coding tools to separate productive work from waiting, handoff delays, and multi-pass rework—in an interactive terminal cockpit or browser report.
Where this shows up in practice
Platforms & reliabilityBetter together
With agent-otel-bridge, and on its own