skills.by true||false
LibraryHow to run

A product of true||false

Every item lists its source, license and what we changed.

library
agentsubagentstrueorfalse

verify

Runs all checks to completion and reports pass/fail per suite with quoted errors, separating code failures from environment problems.

looks for

  • is it green?
  • a failing pre-commit hook or CI job
  • a change about to be committed or merged
  • a work order about to close

outputs

  • → verdict (green or N failing suites)
  • → one line per suite with counts
  • → quoted errors, classified as code / environment / not run
drawing flowchart…
flow.mmd

SKILL.md

You run the project's checks and report results faithfully. You do not fix anything unless asked.

Run

From the repo root, in this order. Keep going after a failure: the caller wants the full picture, not the first error.

{{GENERATED_CHECK}}        # e.g. regenerate in --check mode: no diff allowed
{{TYPECHECK_COMMAND}}      # e.g. npx tsc --noEmit
{{LINT_COMMAND}}           # e.g. npm run lint
{{TEST_COMMAND}}           # e.g. npx vitest run
{{BUILD_COMMAND}}          # e.g. npm run build
{{SMOKE_COMMAND}}          # optional: start server, curl key routes, stop it

Notes:

  • A cold toolchain download is slow, not hung. Wait before concluding.
  • A "check" mode on generated files means regenerating produces no diff. If it reports stale, the fix is to regenerate, not to hand-edit the output.
  • If a test runner reports zero tests, that is a finding, not a pass.

Report

Lead with the verdict: green, or the count of failing suites.

Then one line per suite: name, pass/fail, counts. For each failure quote the actual error (assertion, type error, lint line). Don't paraphrase or summarize it away.

Classify every problem as one of:

  • a genuine failure in the code under test;
  • an environment problem (missing SDK, no network, unset env var);
  • a check you could not run at all, and why.

If you skipped a suite, say so. Never report green for something you did not execute.

Customize for your project

Fill in the commands above (one block per package or module in a monorepo).

audit trail

source
universalized from a private production project's agent setup
author
trueorfalse.dev
license
MIT
imported
2026-11-13
verified
2026-11-13
changes
Removed the project's submodule command list, toolchain paths and contract-test targets; replaced with a fill-in check table and generic notes.
privacy scan
pass
scripts
none
in the zip
1 file

works with

Claude, Codex, ChatGPT

tools: Bash, Read, Grep, Glob · model: sonnet

pairs with

  • personatrueorfalse

    qa-verifier

    Runs the real checks, records an honest PASS/FAIL/PENDING report in the work order, and never fakes green.

    • implementation finished
    • a request to close a work order or merge
    • is it green?
    Claude · Codex · ChatGPT
  • skilltrueorfalse

    debug

    Runs a reproduce → isolate → hypothesize → fix → prove loop so fixes address causes, not symptoms.

    • an error message or stack trace
    • a failing or flaky test
    • it worked yesterday
    Claude · Codex · ChatGPT
  • workflowtrueorfalse

    release

    Maps which systems a change touches, what must move together, and records each shipped version in an immutable release table.

    • "ship it", "cut a release", "tag a version"
    • a change that touches both a server and its clients
    • docs or store listings that may lag the code
    Claude · Codex · ChatGPT
  • workflowtrueorfalse

    work-order

    Turns an ask into a scoped work order file with acceptance criteria, story points and verify commands.

    • a task accepted with no written scope
    • "add X" requests that touch several files or modules
    • scope creep mid-task
    Claude · Codex · ChatGPT