Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/LucasDuys/forgeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/lucasduys/forge/backprop)<a href="https://agentmods.dev/commands/lucasduys/forge/backprop"><img src="https://agentmods.dev/badge/commands/lucasduys/forge/backprop/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/lucasduys/forge/backprop"><img src="https://agentmods.dev/badge/commands/lucasduys/forge/backprop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00012 | $0.00628 |
| Opus 5.5 | $0.00005 | $0.00251 |
| Sonnet 5.5 | $0.00002 | $0.00126 |
| Haiku 4.5 | $0.00001 | $0.00063 |
Grade A, and why
backprop scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Forge Backprop
Trace a runtime bug back to a specification gap, update the spec, generate a regression test, and log the correction for pattern detection.
Pre-flight Check
-
Verify
.forge/exists. If it does not, stop and tell the user:.forge/not found. Run/forge brainstormfirst to generate specifications. -
Verify
.forge/specs/contains at least onespec-*.mdfile. If none found, stop and tell the user:No specs found in
.forge/specs/. Run/forge brainstormfirst to create specifications. Backprop requires existing specs to trace bugs against. -
Ensure
.forge/history/exists. If not, create it:mkdir -p .forge/history
Parse Arguments
Parse from $ARGUMENTS:
| Input | Format | Description |
|---|---|---|
| Bug description | Quoted string (first positional argument) | Natural-language description of the bug |
--from-test PATH |
Path to a test file | Run this test, capture failure output, use it as the bug description |
Exactly one of the two inputs must be provided. If neither is found, stop and tell the user:
Usage:
/forge backprop "description of bug"or/forge backprop --from-test path/to/failing-test
--from-test mode
If --from-test PATH is provided:
- Verify the test file exists. If not, stop and tell the user the file was not found.
- Detect the test runner from the file extension and project config:
.test.ts,.test.js,.spec.ts,.spec.js— look for Jest, Vitest, or Mocha in package.json.test.py,test_*.py— use pytest.test.go— usego test- Other — ask the user how to run it
- Run the test, capturing stdout and stderr.
- Read the test source code.
- Combine the failure output and source code into the bug description context for the skill.
Invoke Backpropagation
Invoke the forge:backpropagation skill with:
- The bug description (either from the quoted argument or from the test failure output)
- The test file path (if
--from-testwas used) - The test source code (if
--from-testwas used) - The list of spec files in
.forge/specs/
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 58 lines · 12 tokens per session scan A c95b8c5676c2
backprop is a command published in the GitHub repository LucasDuys/forge (56 stars, last pushed 3mo ago), licensed MIT. It adds 12 tokens to every session and 628 once invoked, about $0.0000 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
dev-debug
Methodical and systematic bug diagnostic and resolution.
work-flow-bugfix
Complete workflow to fix a bug, from diagnosis to deployment.
eg-fix-bug
Fix a bug using the elephant/goldfish workflow — problem doc, goldfish diagnosis check, failing test, fix, review, validate.
unity-ralph
Relentless verify-fix loop — refuses to stop until the project is clean. Runs unity-verifier repeatedly with configurable max iterations and stall detection.
svelte:test-fix
Troubleshoot and fix failing tests in Svelte/SvelteKit projects, including debugging test issues and resolving common testing problems.
verify
Run verification loop to validate implementation.