Tribunal provides a governance and audit layer for AI coding agents, consolidating logs, policy enforcement, and security monitoring across tools such as Claude Code, Cursor, Copilot CLI, and Codex CLI. It is designed for teams and organizations using multiple AI-powered coding agents, offering a unified audit log that captures prompts, tool calls, file edits, commands, and cost data from each integrated agent. Tribunal normalizes these events into a common schema, enabling users to search and review activity across agents in a single timeline.
The platform features a policy engine that applies YAML-based rules to agent events, supporting actions such as allow, warn, ask, or deny. Pre-built policy packs address common governance needs, including restricting secrets access, preventing production writes, and supporting SOC2 compliance baselines. A prompt-injection scanner is integrated, utilizing multiple regex families and detection for bidirectional and zero-width characters to identify and alert on suspicious input in real time. Tribunal also tracks cost telemetry, monitoring tokens and dollars spent per agent, session, and user, with the ability to set hard caps and receive Slack alerts if spending exceeds limits.
Tribunal is delivered as a local-first daemon that runs on localhost and stores event data in an append-only SQLite log. Users may optionally batch and ship events to a cloud dashboard for team-wide visibility via a Cloudflare Worker, but cloud usage is not required. The platform supports role-based access, retention controls, signed exports, and offers a compliance plan with SAML/OIDC single sign-on and customer-managed encryption keys. Installation is performed via pipx, and adapters are provided for each supported agent to normalize their logs.
The tool is open-source under the MIT license for the OSS tier, which is free and includes the full CLI, local dashboard, adapters, and all shipped policy packs. Paid plans are available: the GitHub Team plan provides a hosted dashboard, team policies, Slack alerts, and GitHub SSO for a monthly per-seat fee, while the Compliance plan offers advanced compliance features and dedicated support for a flat monthly rate. Tribunal positions itself as a single governance solution for organizations managing multiple AI coding agents.
In the AI & ML space, tribunal takes a focused approach. Ensures the quality and security of AI-generated code by enforcing tests and scanning for secrets. It is built as an open-source project for AI developers and coding teams. tribunal is open source under the MIT license. It runs on the web and the command line.
It is developed by thebotclub, and the product first shipped in 2026. Development happens publicly on GitHub with 31 commits in the last 90 days. PulseGate's similarity index finds few close equivalents — tribunal occupies a relatively distinct niche. Key capabilities include TDD enforcement, secret scanning, and audit trails.
Latest indexed changes and source events
tribunal.dev discovered by the PulseGate indexer
Other apps tracked under the same category.
