forge

Writes, debugs, and reviews code with systematic root-cause analysis and verification.

1|Updated Jun 1, 2026
One-click install
npx skills add https://github.com/kridaydave/My_Skills --skill forge-kridaydave
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: forge
Source: https://github.com/kridaydave/My_Skills/tree/main/forge
Command: npx skills add https://github.com/kridaydave/My_Skills --skill forge-kridaydave

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Code that looks right but fails at runtime wastes hours of debugging. This Skill enforces a compile-clean discipline: every snippet is mentally traced, verified against real library APIs, and debugged through systematic root-cause analysis instead of guess-and-check patching. ## Core Features & Use Cases - Verified Code Generation: Produces complete, runnable code with imports, type hints, and edge cases handled, flagging anything unverifiable with explicit # verify: markers. - Systematic Debugging: Applies a reproduce-isolate-hypothesize-disprove loop for intermittent bugs, heisenbugs, flaky tests, and build failures, fixing root causes rather than symptoms. - Structured Code Review: Delivers verdicts with severity-ranked findings (critical, warning, nit) plus honest assessment of what is solid. - Use Case: A test passes alone but fails in the full suite. The Skill guides you to lock test ordering, isolate the failing pair, identify shared-state mutation as the cause, and fix the fixture at the root. ## Quick Start Ask the assistant to debug why your test suite fails intermittently and trace the failure to its root cause.

Frequently Asked Questions about forge

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

FAQPage Schema
How do I debug a test that passes alone but fails in the suite?▼

This is typically order-dependent shared state. Lock the test order with a fixed seed, isolate the minimal pair of tests that triggers the failure, then check for module-level globals, database rows, or temp files mutated by an earlier test. Reset that state in setup to confirm the cause.

How to fix code that looks correct but fails at runtime?▼

Trace the code as the compiler or interpreter would: verify every import exists, check that each library method exists in the installed version, confirm types line up, and ensure every variable is defined before use. AI-generated code often invents plausible but nonexistent API methods.

What languages does this coding skill support?▼

It defines explicit standards for Python (type hints, pathlib, dataclasses), TypeScript (strict typing, async/await), Go (error wrapping, table-driven tests), Rust (Result propagation, no unwrap), and SQL (parameterized queries, no SELECT *).

Why does AI-generated code fail even when it compiles?▼

Compilation only checks syntax, not semantics. Hallucinated library methods are often syntactically valid attribute lookups that only fail at runtime. Verify each method against the installed package version's documentation before trusting the code.

When should I ask a question versus assume during a coding task?▼

Ask one specific question only when the ambiguity changes the architecture. When a reasonable default exists, state the assumption explicitly and deliver the complete solution rather than stalling on clarification.