When a test fails or a bug surfaces, most workflows fix the code and move on. This skill does something different: it traces the root cause, asks whether a new invariant would prevent recurrence, and appends both a bug log entry and a testable rule to your spec. You get a §B row documenting what broke, a §V invariant stating the new constraint, a test that enforces it, and the fix itself, all in one commit. The discipline is simple: bugs should update the specification, not just the implementation. If you're doing specification-driven development or want your project's spec to actually grow smarter over time instead of rotting, this automates the feedback loop that plan-then-execute skips.

skills.sh installs
2.6K
GitHub stars
1.2K
last updated
Aug 21, 2026

Install

$ npx -y skills add juliusbrussee/cavekit --skill backprop --agent claude-code

Installs into .claude/skills of the current project.

Files

backprop — bug → spec

Plan-then-execute fixes the code & forgets. SDD fixes the code AND edits spec so recurrence is impossible. That edit is backprop.

WHEN TO BACKPROP

  • Test failed at /build verification.
  • User reports bug.
  • Post-mortem after production incident.
  • /check flags VIOLATE with root cause found.

SIX STEPS

1. TRACE

Read failure output / bug report. Find exact file:line of wrong behavior. Name root cause in one caveman sentence.

2. ANALYZE

Ask three questions:

  • Would a new §V invariant catch this class of bug? (most common: yes)
  • Is §I wrong — did spec claim shape the code cannot deliver? (sometimes)
  • Is §T wrong — did we build the wrong thing? (rare but real)

3. PROPOSE

Draft the spec change. Never skip §B; §V/§I/§T are case-by-case.

Template:

§B row: B<next>|<date>|<root cause>|V<N>
§V line: V<next>: <testable rule that would have caught it>

Example:

§B row: B3|2026-04-20|refund job ran twice on retry|V7
§V line: V7: ∀ refund → idempotency key check before charge reversal

Sponsors

Details

Repository
github.com/juliusbrussee/cavekit
Categories
Testing & QA, Debugging, Productivity & Planning
First seen
May 16, 2026

More from juliusbrussee/cavekit

All 5
1k-dense-ai avatarscientific-writingk-dense-ai/scientific-agent-skillsDraft, revise, and audit scientific manuscripts or reports with explicit evidence provenance, reporting-guideline coverage, authorship accountability, confident2.6Kskills.sh installs
2amelnagdy avatarwoo-guardamelnagdy/guard-skillsReview generated or changed WooCommerce code — extensions, payment and shipping integrations, checkout customizations, and order/product logic — before it ships2.6Kskills.sh installs
3forcedotcom avatardeveloping-datacloud-code-extensionforcedotcom/sf-skillsDevelop and deploy Data Cloud Code Extensions using SF CLI plugin. Use this skill when creating custom Python transformations for Data Cloud, deploying code ext2.5Kskills.sh installs
4agricidaniel avatarblog-locale-auditagricidaniel/claude-blogAudit a directory of multilingual blog content for completeness, consistency, hreflang correctness, meta-tag parity, and freshness. Builds a translation coverag2.5Kskills.sh installs
5cognitedata avatarskill-creatorcognitedata/builder-skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch for Claude Code or Curs2.5Kskills.sh installs
6nvidia avatarcudaq-guidenvidia/skillsCUDA-Q onboarding guide for installation, test programs, GPU simulation, QPU hardware, and quantum applications.2.5Kskills.sh installs