Business opportunity
Explore tools for building and reviewing software. Test generated changes and keep credentials out of prompts.
These are provider-described capabilities and proposed applications. We have not independently tested the tools or verified business outcomes.
Showing 20 of 20 sourced listings
Alphabetical, not rankedDevelopment
A terminal-based AI pair-programming tool that works with an existing codebase. It includes Git integration and can run configured linters and tests after edits.
Prepare a bounded maintenance patch using an approved model and repository.
Check first: Model access and cost are separate choices, and automatic commits or commands need careful permission and diff review.
Source reviewed
Development
Provides code suggestions and inline development assistance in editors and the command line. Its listed tasks include code changes, documentation, testing and upgrades.
Investigate a maintenance task in an approved AWS-oriented codebase.
Check first: The official page schedules IDE plugin support to end on 30 April 2027. Review the migration path before adopting those plugins for a long-lived project.
Source reviewed
Development
Provides coding-agent workflows for tasks such as pull-request preparation and code review. Its current Cosmos platform coordinates steps from specification through verification.
Prepare a reviewed change proposal from an engineering team’s approved ticket.
Check first: Confirm which current product and access plan fits the task; listed throughput and incident-resolution figures were not verified here.
Source reviewed
Development
Adds codebase context to coding-agent requests and routes them between AI models. Its current product page describes use alongside existing coding tools and gateways.
Evaluate code-context delivery for an approved internal development workflow.
Check first: Routing changes can affect model choice and data handling. Obtain approval for those changes and measure costs rather than assume the advertised savings.
Source reviewed
Development
Builds application projects from prompts and supports iterative changes through chat. Its documentation describes design references, Figma imports and hosting options.
Prepare a small interactive application prototype from an agreed design brief.
Check first: Generated code, authentication, data storage and hosting costs require review before any real-user launch.
Source reviewed
Development
An AI coding assistant that can edit files, run commands and work across a project. Official docs describe terminal, IDE, desktop and web surfaces.
Prepare an isolated bug-fix or test-writing draft in a repository you are authorised to change.
Check first: Most surfaces require an account or supported provider connection. Editorial caution: command execution and file changes need explicit permission boundaries.
Source reviewed
Development
An open-source coding agent with editor, terminal and SDK surfaces. It can plan changes, edit project files and run commands with configurable approval controls.
Prepare a bounded maintenance patch in a repository whose owner has authorised the work.
Check first: Review command permissions and generated diffs, and account for the chosen model connection; an agent’s report is not a passing test.
Source reviewed
Development
Reviews pull requests and presents findings or suggested changes. Its platform also provides review prioritisation and code-health views.
Add a preliminary automated review to an owner-approved software maintenance process.
Check first: Automated findings can miss defects or be incorrect; an accountable reviewer must decide what to change and verify the result.
Source reviewed
Development
Provides AI-assisted code editing and agent tasks across a codebase. Its desktop and command-line surfaces support reviewing generated changes.
Prepare a reviewable prototype or refactoring proposal for an agreed software task.
Check first: Constrain file and command access and inspect the actual diff; an agent’s completion message is not acceptance evidence.
Source reviewed
Development
Works on coding tasks in cloud or local environments and can prepare pull requests. The product also describes testing changes in its own browser.
Prepare a bounded bug-fix proposal and test evidence for a repository owner to assess.
Check first: Review the actual diff and reproduce relevant checks; vendor demonstration sessions are not evidence that your task was completed.
Source reviewed
Development
Factory’s coding agent plans, implements and validates software changes. Its platform provides task monitoring and pull-request review tools.
Prepare a small, testable change from an approved engineering brief.
Check first: Keep command access and release permissions bounded; vendor descriptions of validated software do not replace independent acceptance checks.
Source reviewed
Development
Provides code completion, code explanation and proposed edits in development tools. Its agent features can work on coding tasks for human review.
Prepare a prototype feature or maintenance patch in an approved repository.
Check first: Review and test generated code, check repository permissions and keep production deployment behind a separate approval.
Source reviewed
Development
A JetBrains coding agent with planning, implementation and human approval controls. Its CLI supports project guidelines and different model connections.
Prepare a planned code change with tests and a human-reviewed diff for an existing project.
Check first: Model access, execution permissions and usage limits need separate review; no benchmark or cost claim was tested here.
Source reviewed
Development
An open-source coding-agent platform with local and cloud options. Its engineering workflows include pull-request review and proposed fixes for CI failures.
Investigate a failing test in an approved repository and prepare a candidate patch.
Check first: Restrict repository and integration permissions, and inspect every proposed change; scheduled workflows require separate operational approval.
Source reviewed
Development
Provides AI code review using codebase, pull-request history and requirement context. Its tools support shared engineering rules and local review workflows.
Prepare a review checklist and candidate findings for an approved development team.
Check first: Inferred rules and review comments need human validation; the tool does not certify code quality or compliance.
Source reviewed
Development
Builds and revises app projects through chat. Its platform includes service integrations and automated application-testing features.
Prepare an interactive prototype for a client to evaluate before a separately scoped implementation.
Check first: Generated applications still need independent security, access, billing and release checks; no production-readiness claim was verified.
Source reviewed
Development
Provides AI-supported static application security testing with contextual findings and proposed fixes. It can integrate checks into editors and pull requests.
Provide a preliminary code-security review for a repository whose owner has authorised scanning.
Check first: Review findings and test any proposed fix; a scan does not establish that an application is secure.
Source reviewed
Development
Uses repository context for code chat, code edits and completion. It can draw context from local and remote code through Sourcegraph’s search tools.
Help an authorised engineering team understand and draft changes across a large codebase.
Check first: The reviewed documentation identifies Sourcegraph Enterprise support. Confirm access and repository-data handling before selecting it for a project.
Source reviewed
Development
An open-source, self-hosted AI coding assistant with code completion and question answering. It supports integration with existing development infrastructure.
Evaluate code assistance within an engineering team’s approved hosting environment.
Check first: Self-hosting still needs suitable compute, access controls and maintenance; open-source availability does not establish secure operation.
Source reviewed
Development
Zencoder’s coding app coordinates agents for specification, implementation and verification. Its workflows can run isolated tasks and cross-agent reviews.
Prepare a feature prototype with a written specification and candidate tests for an engineer to review.
Check first: Independently inspect the code and test results; a review by another model does not establish correctness or production readiness.
Source reviewed