According to a GitHub blog post published June 15, 2026, GitHub Copilot CLI now exposes built-in "slash commands" that let developers control the terminal AI agent from the command line. The blog shows users can type "/" to see a scrollable list of available commands and highlights actions such as switching models with /model, inspecting token usage, and resuming past sessions. A report on Let's Data Science lists specific commands including /clear, /add-dir, /cwd, /model, and /session, and frames the controls as improving speed, predictability, and auditability by scoping file access and managing session context. The GitHub post also notes the /model selector displays model capabilities, availability, and cost information. The release targets developer ergonomics for CLI-driven workflows and session management in Copilot CLI.
What happened
According to the GitHub blog post dated June 15, 2026, GitHub Copilot CLI introduces built-in slash commands that act as a command center inside the CLI. The post explains typing / brings up a scrollable list of supported slash commands. The blog highlights using /model to switch models and shows model capabilities, availability, and cost alongside each option. The post also describes in-terminal actions for inspecting changes, managing context, and resuming past sessions.
Technical details
Per the GitHub blog, slash commands expose controls for guiding Copilot behavior, inspecting changes, managing context, and handling permissions. Reporting on Let's Data Science lists concrete commands now available to users: /clear, /add-dir, /cwd, /model, and /session, among others. The blog indicates the command UI surfaces model metadata and token usage information so developers can choose appropriate models from the terminal.
Editorial analysis
Command-driven controls like slash commands consolidate common controls into keyboard-first workflows. Companies and teams that rely on terminal-based developer flows typically benefit from predictable, auditable operations when controls are explicit rather than embedded in natural-language prompts. This pattern reduces ambiguity in file-scope access and session state management, which is relevant for compliance-sensitive environments.
Context and significance
For practitioners: The change is an ergonomics and governance improvement rather than a model architecture update. Developers who already embed Copilot into REPLs or CI-driven flows will find it easier to script repeatable interactions and scope file access without leaving the terminal. Tooling that exposes model metadata and cost inline helps teams compare tradeoffs when choosing lighter-weight or higher-reasoning models during iterative development.
What to watch
Observers should track whether organizations adopt slash commands as part of standardized developer policies, and whether GitHub expands the command set to cover finer-grained permission controls, audit logs, or automation hooks. Also watch for integration with existing CLI tooling and shell automation patterns that could enable reproducible developer-agent sessions.
Scoring Rationale #
This is an incremental but practical tooling update that improves developer ergonomics and governance for CLI-driven Copilot workflows. It is useful for practitioners but not a paradigm-shifting model or infrastructure release.
Practice with real Health & Insurance data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Health & Insurance problems