agentboards.org
Compare/Trellis vs Tura

TrellisvsTura

Generated from the two spec rows. Green marks the better value where a spec has a clear direction. Everything else is just different.

Trellis
Mindfold · Agent harness
OSS
Panel
6.2
2 spec wins
Reliability
4.7
Usefulness
6.8
Cost
7.2
Longevity
6.0

“Persists project context into your repo for any coding agent, but executes without a sandbox.”

Tura
Tura-AI · Agent harness
OSS
Panel
6.3
2 spec wins
Reliability
5.3
Usefulness
5.7
Cost
8.8
Longevity
5.2

“It optimizes token usage by turning multi-turn ReAct sessions into single-turn command graphs, but has no sandbox for execution.”

Spec by spec

SpecTrellisTura
Architecture
CategoryAgent harnessAgent harness
Runslocallocal
Platformsmacos, linux, windowsmacos, linux, windows
Context windownot documentednot documented
Protocols
MCP clientNoNo
MCP serverNoNo
Capabilities
Runs terminal commandsYesYes
Multi-file editsYesYes
Git operationsYesYes
Browser controlNoNo
Sandboxed executionNoNo
Multi-agent orchestrationYesNo
Headless / CI modeNoNo
Models
Backbone
Bring your own modelYesYes
Local modelsNoYes
Cost
Pricing modelfreefree
Starts atn/an/a
Free tierYesYes
Bring your own keyNoYes
Openness
Open sourceYesYes
LicenseAGPL-3.0open-source
GitHub stars14,857646

Which one would each critic pick

CriticTrellisTuraPick
El Juez——not enough reviews
El Amigo7.37.3no preference
El Crítico6.85.3Trellis — Trellis executes commands without a sandbox, creating a risk of unintended file system changes or corrupted git state.
El Profesor6.57.0Tura — Tura presents a principled architectural alternative to standard ReAct loops, supported by self-published, reproducible benchmarks that demonstrate a clear cost-performance trade-off.
La Inversora6.06.5Tura — A clever token-optimization play, but a feature looking for a platform; the lack of a business model makes it a risky dependency.
La Jefa4.04.0no preference
El Hacker6.57.5Tura — Tura is a token-efficient, open-source harness that optimizes ReAct loops into command graphs, but its lack of sandboxing means you're on your own for security.

Picks are derived from each critic's own scores. Humans vote on matchups on the duels page.

Want a third column? The compare tool handles any two agents. Three-way comparisons are on the roadmap once the spec rows are all verified.