htcc-verification-matrix

## Use this skill when - You edited htcc code, tests, CI, scripts, examples, or documentation and need to know what to run. - You are preparing a final report after implementation. ## Do not use this skill when - No files changed and the user only asked for discussion. ## Universal checks - Haskell edits: `stylish-haskell -i <changed .hs files>`. - Haskell edits: `lsp_diagnostics` on each changed `.hs` file when the diagnostics tool is available. If unavailable, state that and substitute `stack build` plus the relevant tests. - Any edit: `git diff --check`. - YAML edits: parse `.yml` / `.yaml` files. If `yaml-ls` is unavailable, use `ruby -e 'require "yaml"; YAML.load_file("path")'` or equivalent. - Before final answer: `git status --short` and report uncommitted files. ## Change-type matrix | Change type | Required tests | Manual QA | |---|---|---| | Parser combinators / AST typing | `stack test --test-arguments components`; full `stack test` for core parser behavior | malformed input through `stack exec htcc -- /dev/stdin`, often with `timeout` | | CLI diagnostics / app/Main.hs | relevant subprocess tests or full `stack test` | run `stack exec htcc -- ...`, capture exit code, stdout, stderr | | Output replacement / permissions | `stack test --test-arguments components`; full `stack test` if behavior is shared | create temp files/dirs with modes, run `htcc -o`, verify output and modes | | CI workflow or Travis | `stack test --test-arguments <added command>`; YAML parse | inspect `git diff` and ensure every explicit list includes intended command | | Docker command behavior | `stack test --test-arguments docker` and `stack test --test-arguments docker --test-arguments --clean` when Docker is available | do not fake Docker success; report unavailable Docker explicitly | | Examples | `cd example && make && cd dist && ./knapsack && ./merge_sorting_linked_list && ./shuffle_and_sort`; `cd example && make docker && make clean_docker` only when Docker is available | execute generated examples when requested or affected | | Documentation only | no Haskell formatter unless code changed; run referenced commands when they are the contract and feasible | check links/commands touched by the edit when practical | ## Evidence standard - A passing unit test is not enough when user-visible CLI behavior changed; run the CLI. - A parser test is not enough for hang bugs; include timeout-guarded manual QA. - Use `timeout` or `gtimeout`, whichever exists on the host, and report the command used. - A successful build is not enough for CI edits; parse YAML and run the newly added command locally when possible. - When Docker is unavailable, substitute Docker checks with non-Docker example/build tests where relevant plus command/diff inspection, then report Docker behavior as residual risk. - If a tool is missing, state the exact missing tool and the substitute check. ## Reporting template ```text Changed: <paths> Verification: <commands and pass/fail> Manual QA: <input, exit code, key stdout/stderr or output-file observation> Review: <reviewer outcomes> Remaining: <none or explicit follow-up> ```