executing-plans-extras

Verifies plan-specified CLI commands produce declared evidence beyond exit status.

1|Updated Dec 13, 2025
One-click install
npx skills add https://github.com/ZK-Theory/TDL --skill executing-plans-extras-zk-theory
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: executing-plans-extras
Source: https://github.com/ZK-Theory/TDL/tree/main/.agents/skills/executing-plans-extras
Command: npx skills add https://github.com/ZK-Theory/TDL --skill executing-plans-extras-zk-theory

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? A zero exit status only proves a process terminated normally, not that the intended handler actually ran. This Skill closes that gap when implementation plans specify exact CLI, module, migration, or validation commands whose success depends on observable output or state changes. ## Core Features & Use Cases - Exact-Command Verification: Runs the literal command from the plan in the specified directory and environment, never substituting a unit-level function call. - Evidence-Based Assertions: Declares required evidence before execution (stdout/stderr, created artifacts, database state, exit code) and adds subprocess-level tests asserting it. - No-Op Detection: Treats exit zero with missing output or missing state change as a failed verification, investigating entrypoints like main, argument dispatch, and early returns. - Use Case: A plan requires running a database migration command. The command exits 0 but no migration applied because of a broken argument dispatch. This Skill catches the no-op by asserting the declared database state change. ## Quick Start Use executing-plans-extras to verify that each CLI command in my implementation plan actually produced its declared evidence before marking the task complete.

Frequently Asked Questions about executing-plans-extras

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I verify a CLI command actually ran instead of exiting silently?

Run the literal command from the specified directory and environment, then assert its declared evidence such as expected stdout, created artifacts, or database state changes. Exit code zero alone does not prove the intended handler executed.

Why does my command exit 0 but do nothing?

Common causes include a missing or broken __main__ entrypoint, incorrect argument dispatch, or an early return that skips the handler. Investigate these entrypoints and add a subprocess-level test asserting the expected observable output or state change.

Can a unit test replace running the actual CLI command?

No. A direct handler unit test can supplement verification but cannot replace it, because it bypasses entrypoints, argument parsing, and dispatch logic where no-op failures typically occur. Always run the literal command end to end.

When should I use exact-command verification during plan execution?

Use it whenever an implementation plan names exact CLI, module, migration, or validation commands, especially when exit status alone could hide a no-op entrypoint or when completion depends on command output or a state transition.

What evidence should be recorded for each verified command?

Record the literal command, its exit code, the captured evidence such as stdout or changed artifacts, and the negative or no-op check result in the task report before marking the task complete.