Write deterministic git-spice test scripts with txtar data blocks and golden-file comparisons.
npx skills add https://github.com/abhinav/git-spice --skill test-script
Please help me install this Agent Skill. Skill: test-script Source: https://github.com/abhinav/git-spice/tree/main/.claude/skills/test-script Command: npx skills add https://github.com/abhinav/git-spice --skill test-script
This guide helps teams create and maintain deterministic test scripts for git-spice workflows, enabling repeatable end-to-end validations of branch stacking, ShamHub forge simulations, and golden-file comparisons.
Create a new test script under testdata/script named for the scenario and run it with mise run test:script --run <name>.
High-intent search queries and answers about installing and using this skill.
Deterministic test scripts for git-spice workflows use txtar-style files with structured data blocks to specify fixed time and author values, enabling repeatable end-to-end validation of branch stacking operations. You place these plain-text scripts under testdata/script/ to ensure reproducible outputs.
Simulating forge interactions when testing git-spice is achieved through ShamHub forge simulations, which allow you to validate interactive prompts and stack operations against a simulated remote repository deterministically within your test scripts.
Golden-file comparisons for txtar test data are run and updated using the test harness via the command mise run test:script --run <name>, which executes the scenario and validates deterministic outputs against the expected golden files.
Yes, this testing approach validates interactive prompts in git-spice by embedding prompt interactions within structured txtar-style test scripts, ensuring reproducible validation of stack operations and forge interactions without manual input.
Git-spice end-to-end tests fail with non-deterministic outputs when scripts omit explicit time and author specifications. Applying deterministic test data blocks via txtar formats enforces fixed commit metadata, resolving golden-file comparison mismatches.
Yes, you can use txtar test scripts for end-to-end validation of stack operations by defining plain-text test data blocks under testdata/script/ and executing them with the mise test harness to compare deterministic outputs.
chadanandaCreate and validate isolated Git worktrees for feature work.
facebookRun and debug Pysa taint analysis integration tests with expected output updates.
coval-aiConfigure and execute AI evaluation workflows on the Coval platform.
maxculen86Test Bubbletea Go applications with table-driven, model, teatest, and golden-file patterns.
