Done-Check: Tests, Leftovers & Junk Files

    1

    Stops "done" without evidence: runs the tests and flags debug prints, skipped tests and junk files in one call.

    Free

    1 useSecurity scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Done-Check: Tests, Leftovers & Junk Files

    Example session with this skill installed

    I've updated slugify to handle punctuation. Verify everything before I commit.

    • Read your context and instructions
    • Compiled the done-check

    Real session on a small Python repo. The script output is pasted as-is.

    Summary

    The first check found 3 problems: a debug print, a skipped test and a .orig backup file. After fixing them, the second check passes with 2 tests run and nothing skipped. The behavior the request asked for is confirmed by running it.

    1. First check

    $ python3 scripts/verify.py
    [PASS] tests: PASS (exit 0) 1 passed, 1 skipped in 0.07s
    [info] changed files (3): tests/test_text.py, text.py, text.py.orig
    [FAIL] no leftovers in the diff (3 found)
        tests/test_text.py: test skipped: @unittest.skip("fix later")
        text.py: debug print added: print("DEBUG slugify input:", s)
        text.py.orig: debug print added: print("DEBUG slugify input:", s)
    [FAIL] no junk files: text.py.orig
    VERDICT: NOT ready — fix the FAIL items first
    

    The tests "passed" only because the new punctuation test was skipped.

    2. Fixes

    ItemFix
    print("DEBUG slugify input:", s) in text.pyRemoved
    @unittest.skip("fix later") on test_punctuationRemoved, so the test now runs
    text.py.origDeleted

    3. Second check

    $ python3 scripts/verify.py
    [PASS] tests: PASS (exit 0) 2 passed in 0.05s
    [info] changed files (2): tests/test_text.py, text.py
    [PASS] no leftovers in the diff
    [PASS] no junk files
    VERDICT: ready — re-read the request once more, then finish
    

    4. Request vs. evidence

    RequirementEvidence
    slugify handles punctuationtest_punctuation runs and passes
    Existing behavior unchangedtest_basic passes
    $ python3 -c 'from text import slugify; print(slugify("Hi, there!"))'
    hi-there
    

    Ready to commit: 2 files changed (text.py, tests/test_text.py), 2 of 2 tests passing, no leftovers.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Code quality often suffers from "ready to ship" assumptions that ignore side effects. This skill forces a systematic pre-flight check to ensure the work actually meets the original requirements without leaving behind technical debt. It stops the common habit of reporting success before verifying the entire suite or cleaning up the workspace.

    What it does

    • Automated Scanning Identifies debug prints, breakpoints, and merge conflict markers in the diff.
    • Requirement Mapping Has the agent re-read the request and point each requirement to a file or test.
    • Environment Cleanup Detects junk files like .orig or scratch files that shouldn't be committed.
    • Test Run Runs the detected test command and summarizes failures and asserts.
    • CLI Validation Has the agent run the commands the request mentions and compare the output.

    How it works

    1. Run scripts/verify.py to scan for leftovers and execute the project's test command.
    2. Address every FAIL item in the generated checklist, including skipped tests and junk files.
    3. Map every requirement from the original prompt to a line of code or a specific test case.
    4. Execute any CLI examples provided in the request to ensure the output matches expected results exactly.

    Frameworks & tools

    This tool uses Python 3 standard library scripts for environment-agnostic verification. It integrates with existing test runners like pytest or unittest.

    Why this beats prompting it yourself

    Generic prompts often fail to catch "invisible" issues like .orig files or leftover print() statements. This skill provides a structured Python-backed checklist that ensures every claim of success is backed by empirical evidence rather than a guess.

    Use cases

    • Preventing debug code from reaching production branches.
    • Ensuring refactored code doesn't break existing features in the suite.
    • Verifying CLI tool outputs against strict formatting requirements.
    • Catching .orig, backup and scratch files before a commit.

    Known limitations

    Requires a local Python 3 environment. Test commands must be provided manually if the project uses non-standard test runners.

    Free and open source (Apache-2.0), by the author of rosetta-js, a terminal coding agent for open-weight models.

    Requires Python 3 and a git repository. Auto-detects pytest/unittest, npm test, go test and cargo test; other runners need --test "CMD". The leftover scan is pattern-based, so it can miss unusual debug code.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    1 use

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 1 month ago

    • Free with an Agensi account

    Requires

    python3
    git

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Environment Variables

    Shell runs verify.py and the project tests. Read Files: it scans the git diff and working tree for leftovers. It never edits or deletes anything itself; the agent fixes the flagged items. Environment variables are only passed through to the test command. No network access.

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    This skill acts as an automated "pre-flight" gatekeeper, scanning your diffs for forgotten debug prints and temporary files while mapping original requirements directly to passing test cases to prevent premature completion reports.

    Any agent that supports SKILL.md and can run shell commands, for example Claude Code, Codex CLI, Cursor or Gemini CLI. It needs Python 3 and git.

    The SKILL.md workflow, two Python 3 standard-library scripts (verify.py and test_summary.py), a smoke test that checks them, and the Apache-2.0 license. It's free.

    Yes. The workflow has the agent run the exact commands the request mentions and compare the output with what was asked. The comparison is done by the agent; the scripts cover tests, leftovers and junk files.

    No configuration. verify.py picks the test command itself (pytest or unittest, npm test, go test, cargo test). Pass --test "CMD" only for anything else. It needs a git repository because it scans the diff.

    The skill currently requires a local Python 3 environment and performs best with standard test runners; if you use a proprietary or highly non-standard testing tool, you will need to provide the execution command manually.

    New in Code Quality & Review