Is the Backpropagation skill free?

Yes. The skill itself is free from LucasDuys/forge. SkillProof publishes the install command and an independent test verdict at no cost.

What is the Backpropagation SkillProof Score?

9.2/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 5/5.

How do I install Backpropagation?

Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.

Can I use Backpropagation with Cursor, Copilot, Gemini CLI, Codex or other AI tools?

The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.

Backpropagation

Trace a runtime bug back to the spec gap, close it, and generate a fail-first regression test

by LucasDuys · LucasDuys/forge

Tested · Works

Install — copy the command Download source ↗

Test report

Verdict: Tested · Works
Score: 9.2/10
Tested: Jul 18, 2026
Environment: Claude Code 2.x (agent harness)
Upstream re-checked: Jul 30, 2026 · fd05807

⚠ This skill is no longer available upstream. Our re-check on Aug 10, 2026 couldn't find it any more (repo unreachable/deleted). The test below is what we measured on Jul 18, 2026 and we're leaving it up as a record — but there is nothing left to install, so we've removed the command.

Built a real scenario (spec R001 'user registration' + a register() that let 'Alice@x.com' and 'alice@x.com' both through) and ran the skill's loop end-to-end: it traced to a missing_criterion, proposed a testable duplicate-casing criterion, and the regression test it specified failed against the buggy code (exit 1) then passed after the case-normalization fix (exit 0). A no-skill baseline just patches toLowerCase and moves on, leaving no spec criterion, no durable test, and no pattern log to catch the next casing/whitespace variant.

Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works

Installs cleanly 5/5
Triggers reliably 5/5
Output vs. baseline 8/10
Docs & honesty 5/5

What Backpropagation does

A spec-driven debugging loop: given a bug or failing test, it traces the failure to the missing/incomplete acceptance criterion in .forge/specs, classifies the gap, proposes a spec update (with your approval), writes a minimal regression test proven to fail before the fix, and logs the correction so recurring pattern categories surface systemic fixes. Triggers on 'trace this bug to the spec', 'backprop this failure', 'generate a regression test and update the spec'.

How to install Backpropagation

Nothing to install: the source repository no longer has this skill. If the author brings it back, our daily re-check will pick it up and the command will reappear here.

Commands — how to trigger Backpropagation

/backpropagation Trace a runtime bug back to the spec gap, close it, and generate a fail-first regression test

It also activates on plain-language prompts like these:

Trace this registration bug back to the spec gap that caused it.
Our spec missed this duplicate-email edge case, can you backprop it?
Generate a regression test and update the spec for this failure.

Frequently asked questions

Is the Backpropagation skill free?: Yes. The skill itself is free from LucasDuys/forge. SkillProof publishes the install command and an independent test verdict at no cost.
Does Backpropagation work with Claude Code?: We tested it with Claude Code 2.x (agent harness) on Jul 18, 2026. Verdict: Tested · Works. Built a real scenario (spec R001 'user registration' + a register() that let 'Alice@x.com' and 'alice@x.com' both through) and ran the skill's loop end-to-end: it traced to a missing_criterion, proposed a testable duplicate-casing criterion, and the regression test it specified failed against the buggy code (exit 1) then passed after the case-normalization fix (exit 0). A no-skill baseline just patches toLowerCase and moves on, leaving no spec criterion, no durable test, and no pattern log to catch the next casing/whitespace variant.
What is the Backpropagation SkillProof Score?: 9.2/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 5/5.
How do I install Backpropagation?: Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
Can I use Backpropagation with Cursor, Copilot, Gemini CLI, Codex or other AI tools?: The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.