Skip to content

[deps-release-notes] copilot-cli v1.0.49 action items #641

@github-actions

Description

@github-actions

Release Notes Action Items for copilot-cli 1.0.481.0.49

This issue summarizes upstream release notes for the copilot-cli dependency between the previously pinned version (1.0.48) and the new pinned version (1.0.49), highlighting items that may need follow-up in ado-aw.

The companion version-bump PR is titled chore(deps): update COPILOT_CLI_VERSION to 1.0.49.

Releases analyzed

Notable features for ado-aw to adopt

  • postToolUse hook additionalContext injected as system message: Previously silently discarded; now injected as a system message for the model. If ado-aw pipelines configure postToolUse hooks, this behaviour change may affect agent output. (v1.0.49)
  • Alpine Linux (musl libc) support: Copilot CLI can now run on Alpine Linux. This may open up new base-image options for AWF agent containers. (v1.0.49)
  • COPILOT_PLUGIN_DIR_ONLY env var: Disables automatic plugin discovery, enabling deterministic plugin sets when using --plugin-dir. Useful for ado-aw pipelines that pin specific plugins for reproducibility. (v1.0.49)
  • Hooks fire correctly for sub-agent tool calls: preToolUse, postToolUse, subagentStart, subagentStop hooks now fire for sub-agent tool calls, not just top-level agent calls. ado-aw pipelines using hooks for auditing/safe-output interception should verify behaviour is as expected. (v1.0.49)
  • --plugin-dir and --additional-mcp-config work in --server/--headless mode: Previously these options may not have worked in headless mode. ado-aw runs Copilot CLI in headless/prompt mode; this fix may unlock new MCP plugin configurations. (v1.0.49)
  • auth.redirectPort config option for MCP OAuth: Pins the OAuth callback to a fixed port for MCP servers using static OAuth clients. May be useful for ado-aw MCP server configurations in network-isolated environments. (v1.0.49)
  • Prompt mode (-p) automatically loads workspace MCP sources when folder is trusted: Improves MCP source discovery in prompt mode. Relevant for ado-aw's use of -p/prompt mode. (v1.0.49)
  • Repo hooks in .github/hooks/ now load in prompt mode (-p): ado-aw could leverage this to ship per-repo hook configurations alongside compiled pipelines. (v1.0.49)
  • "None" reasoning effort option: Adds the ability to disable model reasoning in the reasoning effort picker. ado-aw's engine configuration (engine.md) could expose this as a new value for the reasoning-effort field. (v1.0.49)

This issue was opened automatically by the dependency version updater workflow.

Generated by Dependency Version Updater · ● 9.5M ·

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions