Coding agents
Coding and software-engineering agents with source-linked identity and evidence status.
Discover · compare · verify
Discover coding, research, automation, voice and business AI agents using verified primary-source evidence, transparent unknowns and dated comparisons.
Search verified entities and reviewed authority guides.
Explore by category
Categories organize verified records by workflow. They do not imply performance, compliance or ranking.
Coding and software-engineering agents with source-linked identity and evidence status.
Voice-agent platforms. Pricing and latency are unknown unless directly sourced.
Agent frameworks and orchestration systems.
Agent platforms used in business workflows. No pay-to-rank ordering.
AI-agent tooling around CRM workflows; field-level verification is required for product claims.
Customer-support agent tooling with source-linked records.
Sales workflow agent tooling. Results and ROI are not estimated without evidence.
Marketing workflow tooling. No fabricated performance claims.
Research and browsing agents with source-linked identity.
Workflow automation and agent orchestration tools.
AI agents applied to finance workflows — bookkeeping, invoicing, reconciliation and reporting. No fabricated ROI or compliance claims.
AI agents with publicly available, licensed source code. Listed by verified repository identity, not by license claim alone.
AI agents applied to graphic design, visual and creative asset generation. No fabricated licensing or usage-rights claims.
Featured verified agents
Identity evidence is verified independently from pricing, benchmarks, security and compliance claims.
Open-source software-agent project; this site verifies repository identity separately from editorial capability summaries.
Open-source coding-agent project with a public upstream repository.
Open-source terminal coding project with a public upstream repository.
Google-maintained terminal agent project with a public upstream repository.
Qwen-maintained coding-agent project with a public upstream repository.
Software-engineering agent research project with a public upstream repository.
Authority guides
These pages preserve real Search Console demand with substantive canonical answers and dated evidence where the subject is volatile.
Built in India
India relevance comes from primary publisher evidence. It is never converted into an automatic sovereignty or compliance badge.
A Mixture-of-Experts text-generation model with a strong India-language focus. Model-card-reported benchmark results are not promoted as independent rankings.
Large Sarvam model with a first-party model card and downloadable artifacts; publisher benchmarks remain labeled as publisher-reported.
Base model focused on Hindi, English and Hinglish; the source explicitly notes it is a base model rather than a ready-made chat assistant.
Hindi instruction-tuned model built from OpenHathi and the IndicInstruct dataset.
Instruction-tuned 12B model built on Mistral-NeMo architecture with an India/Indic focus.
Open-source model/tooling project with a public, non-archived upstream repository (suno-ai/bark). Capability and benchmark claims are not inherited from repository identity and remain unknown unless separately evidenced.
Developer infrastructure
Frameworks and protocol infrastructure stay distinct from end-user agents so evidence is attached to the right object.
Stateful agent-workflow framework with a public upstream repository.
Open-source application and agent framework with a public upstream repository.
Microsoft-maintained agent framework with a public upstream repository.
Microsoft agent framework retained as a distinct entity; current lifecycle guidance should be read from its upstream source.
Open-source multi-agent orchestration framework with a public upstream repository.
Protocol hub
Source-led MCP guidance. Legacy server detail routes stay noindex until canonical upstream and primary evidence qualify them.
Machine-readable data
Every export below is filtered through the same publication gate as the HTML pages — nothing appears here that isn't also indexable on the site.
The full public catalog — every verified entity across agents, models, frameworks, providers and MCP servers.
Per-type slices of the same gated catalog data.
Plain-text discovery surfaces built for AI crawlers, listing only indexable pages.
Search the same verified entity graph shown on the site, as JSON.
The single canonical sitemap — every URL here is indexable; nothing noindex is ever listed.
The publication rule these exports and the HTML site both obey, explained.
Trust layer
Discovery → normalization → evidence → verification → publication. Donor keywords and old URLs are useful research inputs, but unresolved candidates remain quarantined.