Skip to content

ericrisco/rsc-harness / skills/verify

verify

Verified

Use when work is about to be called done — the ASDD evidence gate, in two modes: per task inside `implement`'s loop (tests, build, security, traceability) and at closing over the whole branch against main (stack scripts/verify.sh, every done-check and acceptance criterion, rollback plan, dated verdict). A check that did not execute is reported not run, never passed. NOT judging the diff by eye (that is `review`, spec-less `code-review`), NOT diagnosis (that is `debug`).

FileKindSize
SKILL.mdskill md21.6 KB
evals/README.mdmarkdown4.2 KB
evals/cases.yamlconfig8.3 KB
references/generic-gate.mdmarkdown2.2 KB

references/generic-gate.md

View on GitHub
# Generic gate — when no stack skill ships `scripts/verify.sh`

A project in a stack the catalog has no gate for (a plain Node CLI, a Rust crate, a shell tool) is
still verified. The gate is the commands the project itself declares, as `sdd-init` recorded them.
This is a first-class path, not a fallback to apologise for.

## Where the commands come from

`02-DOCS/wiki/sdd/config.yaml` → `testing.commands.verify` (and `testing.commands.apply` for the
suite), filled by `sdd-init` from the manifests (`package.json` scripts, `Makefile`, `Cargo.toml`,
`pyproject.toml`…). No config → run `sdd-init` first; do not pick commands from memory.

## The five layers

| Layer | What counts | Typical source |
| --- | --- | --- |
| test | the full suite | `npm test`, `cargo test`, `pytest` |
| lint | the project's linter | `npm run lint`, `cargo clippy` |
| type | the type checker | `tsc --noEmit`, `mypy` |
| build | the build or compile | `npm run build`, `cargo build` |
| audit | dependency vulnerabilities | `npm audit`, `cargo audit`, `pip-audit` |

## Procedure

1. List the five layers. For each, take the command from `config.yaml`; if it declares none, look
   for one in the manifest and say where you found it.
2. Run each command once, after the last edit, from the project root. Capture the exit code and the
   relevant output verbatim.
3. Record every layer as exactly one of:
   - **pass** — the command ran and exited 0;
   - **fail** — the command ran and exited non-zero (a hard FAIL of the verdict);
   - **not run** — no command is declared or the tool is missing. Label it as `verify` does:
     `NO-APLICA` (no such surface, e.g. no types in plain JS), `HERRAMIENTA-AUSENTE` (the surface
     exists, nothing checked it).
4. Write the layers under `## Stack gate` in the record, as `generic gate (config.yaml) → …`, one
   line per layer, with the command that ran.

## Rules that still apply

- A `not run` layer whose surface exists fails the verdict, exactly like a SKIP in a stack script.
- A command the project wrote (a custom check script) is a home-grown gate: watch it fail and pass
  before trusting it.
- Suggest a stack skill if one fits — but the generic gate is the verdict for this run.
rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first.Stars: 194kUpdated Claude CodeCodexVerifiedInstall