Find The Right Tool For
Every Part Of Your Software Factory

Serves Slack, Confluence, Jira and PR history to coding agents over MCP.

Semantic code retrieval across many repositories for MCP clients such as Claude Code, Cursor and Zed.

Cross-repository code search that agents reach over MCP.

Indexes repositories, documentation sites and PDFs and adds a research agent.

Serves version-specific library documentation to coding agents.

Curated repository files kept in sync and reachable through the GitHub MCP server.

Managed organisation CLAUDE.md, .claude/rules/ and auto memory for static conventions.

Open framework of context providers for coding assistants.

Temporal knowledge graph memory built from episodes, with bitemporal validity and hybrid retrieval.

Memory API (remember, recall, improve, forget) that runs on Postgres with pgvector, plus an MCP server.

Memory API with time-aware retrieval, available as open source and as a hosted service.

Memory for users and peers on Postgres with pgvector, hybrid retrieval and a Claude Code plugin.

Agent runtime with self-editing memory blocks.

Memory layer built from memory cubes, with an asynchronous scheduler.

Memory API with a graph and a Claude Code integration.

Single-user memory store that keeps notes as Markdown files.

Library for extracting and consolidating long-term memory in LangGraph agents.

Research library for agentic memory that links and updates notes.

Single Rust binary on Postgres with ontology packs, bitemporal facts, entity resolution and one MCP server per base.

Graph-based retrieval that can keep all its stores in Postgres, with incremental updates and deletes.

Builds community summaries over a graph to answer questions about a whole corpus.

Schema-guided extraction of a knowledge graph from documents into Neo4j.

Small, hackable implementation of the GraphRAG approach.

Vector similarity search inside Postgres, with iterative index scans for filtered queries.

BM25 full-text search inside Postgres, combinable with pgvector through reciprocal rank fusion.

StreamingDiskANN index and binary quantisation for pgvector.

Dedicated vector database with strong support for filtered queries.

Vector database with native hybrid search.

Distributed vector database with native hybrid search.

Search and vector engine with native hybrid ranking.

Search engine with vector and keyword hybrid queries.

Embedded vector database that stores data in files or object storage.

Serverless vector and full-text search on object storage.

Open format for packaging agent skills, supported by more than 30 tools.

Installs skills from an owner/repo on Git into coding agents.

Official registry for MCP servers, self-hostable as a private sub-registry.

Registry of containerised MCP servers, with support for custom organisation catalogs.

Hosted registry and gateway for MCP servers.

MCP gateway with per-tool access control.

Registry of skills and context for coding agents.

Assign an issue and get a draft PR, with fixes for failing Actions runs and merge conflicts.

Runs Claude Code on GitHub from issues and PRs; the web version can auto-fix failing CI and review comments.

Open agent platform where a label or mention on an issue opens a PR.

Label-triggered plan, code, test, self-review and PR flow built on LangGraph.

Picks up backlog issues and opens PRs with the coding agent of your choice.

Cloud coding agent that turns issues and PR mentions into pull requests.

Asynchronous coding agent that plans and opens PRs from GitHub issues.

Autonomous coding agent that works from issues and opens PRs after a plan step.

Cloud agents that take issue or PR triggers and push their work as PRs.

Coding agents that run from issues and PRs, with a planning step.

Cloud development environments and agents, available as an enterprise VPC deployment.

Coding agent for GitHub that opens PRs from issues and mentions.

Minimal agent core to embed in your own tooling.

Markdown-defined agents on Actions that label, close duplicates and comment, with controlled write outputs.

Suggests likely duplicates on new issues; agents can set issue fields over MCP.

Actions that label issues and ask for missing information using GitHub Models.

Suggests duplicates, labels and assignees in Linear, with an agent API.

GitHub Action that labels issues by matching regular expressions.

Durable task engine on Postgres with a TypeScript SDK, live run UI and concurrency limits.

Durable execution library that checkpoints workflows in your existing Postgres.

Durable workflow engine with signals and updates for long waits.

YAML-defined workflows with pause tasks and approval steps.

Workflow and script platform with approval steps.

TypeScript background jobs with waitpoints and realtime streaming.

Step functions with waitForEvent for human or external waits.

TypeScript toolkit for multi-agent networks that run on Inngest.

Durable execution runtime for services and workflows.

TypeScript framework for agents and workflows.

Kubernetes-native workflows defined in YAML, one pod per step.

Workflow orchestration engine with JSON-defined flows and human tasks.

Framework for durable agents built on Dapr workflows.

Python workflow orchestration with scheduling, retries and observability.

Orchestrator for data and AI pipelines, built around software-defined assets.

Python framework for teams of role-playing agents.

Node job queue on Postgres with retries, dead-letter queues, cron and publish and subscribe.

Job queue for Postgres using LISTEN/NOTIFY, with crontab scheduling.

Postgres extension with SQS-style queue semantics and an archive table.

Workflow engine that runs on Postgres and Supabase.

Job queue for Go on Postgres.

Job processing for Elixir on Postgres.

Database-backed job queue for Ruby on Rails.

Messaging with durable consumers per subscriber.

Kubernetes resources for sandboxes, templates, claims and warm pools, with gVisor or Kata.

Kubernetes resources for agents and tool servers, with MCP and OpenTelemetry.

Firecracker sandboxes for agents, self-hostable or run in your own cloud.

Serverless compute with sandboxes for running agent code.

Managed microVM sandboxes, with a bring-your-own-cloud option.

Isolated code execution environments on Cloudflare.

Ephemeral microVMs for running untrusted or AI-generated code.

Persistent sandbox VMs from Fly.io for agents.

Managed sandboxes and agent hosting.

Self-hosted development workspaces on Kubernetes and other infrastructure.

Open-source PR reviewer with /review, /improve and /ask commands, run with your own LLM keys.

AI code review on PRs with chat replies; self-hosting is an enterprise option.

AI code review with whole-codebase context; Helm deployment on enterprise plans.

Review comments from Copilot, native to GitHub pull requests.

PR review that can push autofix commits to the PR.

Multi-agent review of pull requests that posts inline comments.

AI review built into the Graphite stacked-PR workflow.

AI code review and PR summaries.

Code review of GitHub pull requests by OpenAI Codex.

AI code review and refactoring suggestions.

Requirements as text with a traceability graph, source links and JUnit or gcov import.

Requirements in RST or MyST, linked from code comments.

Markdown requirement IDs with coverage tags and gap reports.

Requirements as YAML items in Git, with link validation.

Traces requirements to Git tests and results, with an MCP server for the traceability matrix.

Enterprise application lifecycle management with requirements traceability.

Per-test coverage maps used to select which tests to run.

Test selection driven by which tests cover which code.

Test impact and coverage analytics.

Automated test selection from per-test coverage; currently in beta.

Spec-driven development toolkit that moves from spec to plan to tasks.

Spec, design and task documents with EARS-style requirements.

Merges when required checks and reviews pass, with a queue that keeps PRs up to date.

Rules on age, labels and approvals, with a merge queue and auto-rebase.

Merges on an automerge label and keeps the branch updated.

Merges on label, checks and reviews and updates branches, configured in YAML.

Merge queue with stack-aware merging.

Stack-aware merge queue bundled with Graphite.

Merge queue with batching and stack support.

Merges dependency PRs, with minimumReleaseAge and rebaseWhen options.

Developer portal framework with a software catalog and a plugin system.

Internal developer portal with an agent catalog.

Internal developer portal with scorecards and service catalog.

LLM proxy with per-key, team and user budgets.

Self-hostable LLM traces with cost tracking.

Usage metering and billing infrastructure.

AI gateway with routing, guardrails and observability.

LLM observability and gateway.

Pytest-style LLM evals with golden synthesis and contextual relevancy and recall metrics.

RAG evaluation metrics and test-set generation from documents.

YAML-configured evals and red-teaming with context assertions.

Self-hostable traces, datasets and judge-based evals.

Tracing and RAG relevance evals, free to self-host.

Hosted evaluation and observability platform.

Hosted tracing and evaluation from LangChain; self-hosting is an enterprise option.

Engineering insights with daily per-team Slack notifications.

Engineering metrics platform with daily digests.

AI daily or weekly reports per repository or contributor, sent to Slack.

Developer productivity measurement platform.

Creates a Slack channel per pull request, with daily recaps.

Sends open review requests to Slack on a schedule.

GitHub Action that summarises commits with an LLM and posts to Slack on a cron.

GitHub Action that posts a weekly LLM summary of commits.

Deterministic changelog generator driven by conventional commits.

Automates release PRs and changelogs from commit messages.

Versioning and changelog management based on change files.