LIVE · TAO
TAO$— SUBNETS— VALIDATORS256
Bittensor intelligence updates
Home / AI News/ SN62’s Ridgeline to Clear Your…
AI NEWS

SN62’s Ridgeline to Clear Your GitHub Bug Backlog for ~$1 a Fix

Ridgeline by Ridges (SN62) turns GitHub issues into pull requests for about $1 per coding job. Autonomous AI agents write and test fixes before devs review them.

SN62’s Ridgeline to Clear Your GitHub Bug Backlog for ~$1 a Fix

A well-written GitHub issue can describe a routine bug down to the line of code and still wait weeks for a fix.

Ridges has launched Ridgeline to take on that work. The service assigns an issue to an autonomous coding agent, which works in an isolated environment and submits a pull request.

Ridgeline’s Website

Ridges (SN62) is the subnet that supplies agents. On Ridges, miners compete to build agents that can solve software engineering tasks they have never seen.

SN62 evaluates their work in isolated environments, including applications that run across multiple containers, and takes execution cost into account.

Ridgeline uses agents that perform best in those evaluations.

Ridges’ Website

Each job starts with a fresh checkout. The agent installs dependencies and runs the existing tests to establish which ones pass before any code changes.

It then studies the repository and turns the issue into an implementation plan. Existing test failures stay outside the scope unless resolving them is part of the assignment.

How Ridgeline Agents Work

After implementing the change, the agent runs the most relevant tests first and checks for regressions against the baseline.

A separate review pass looks for incomplete work, changes outside the scope of the issue, and leftover debugging code. The resulting pull request goes to the engineering team, which decides whether to merge it.

How Ridgeline Works

The issue serves as the full brief. Clear acceptance criteria tell the agent what the finished change should do, while details such as affected modules or a requirement to preserve an API help define the scope.

Ridgeline currently verifies Python changes most thoroughly, though agents can read and edit repositories in other languages.

A coding job costs one credit, and a pull request review costs 0.1 credit. A subscription costs $9.99 a month for 10 credits.

Users can earn one credit every three days per 1,000 SN62 alpha they lock. Locked stake continues to earn rewards, and credits earned through staking expire after two weeks.

Ridgeline’s Pricing Arrangement

Ridgeline also supports x402 payments and an MCP integration that lets other agents submit and pay for jobs.

➛ Visit Ridgeline, read its docs, and explore our earlier coverage below:

Enjoyed this article? Join our newsletter

Get the latest TAO & Bittensor news straight to your inbox.

We respect your privacy. Unsubscribe anytime.

The Daily Dispatch

Enjoyed this article?
Join our newsletter

Get the latest TAO & Bittensor news straight to your inbox — every morning before markets open.

IA
Ige A
Editor-in-Chief

No comments yet — be the first.

Leave a Reply