From 1acb8123aec208a09d87f48e386c8d2acad6e9d2 Mon Sep 17 00:00:00 2001 From: dave Date: Tue, 7 Apr 2026 14:16:19 +0000 Subject: [PATCH] huskies: create 494_story_mcp_tool_to_run_project_test_suite --- ...tory_mcp_tool_to_run_project_test_suite.md | 21 +++++++++++++++++++ 1 file changed, 21 insertions(+) create mode 100644 .huskies/work/1_backlog/494_story_mcp_tool_to_run_project_test_suite.md diff --git a/.huskies/work/1_backlog/494_story_mcp_tool_to_run_project_test_suite.md b/.huskies/work/1_backlog/494_story_mcp_tool_to_run_project_test_suite.md new file mode 100644 index 00000000..4374870b --- /dev/null +++ b/.huskies/work/1_backlog/494_story_mcp_tool_to_run_project_test_suite.md @@ -0,0 +1,21 @@ +--- +name: "MCP tool to run project test suite" +--- + +# Story 494: MCP tool to run project test suite + +## User Story + +As an LLM agent or web UI user, I want an MCP tool that runs the project's test suite so I can verify code changes without shelling out to Bash. + +## Acceptance Criteria + +- [ ] New MCP tool `run_tests` that executes `script/test` and returns pass/fail with output +- [ ] Available as a bot command (`test`) in all chat transports +- [ ] Available as a slash command (`/test`) in the web UI +- [ ] Returns structured result: pass/fail, test count, and truncated output for failures +- [ ] Runs in the project root (or optionally in a specified worktree path) + +## Out of Scope + +- TBD