Verify Supervise project rules against codebases and emit rule-application reports.
npx skills add https://github.com/vTRKA/supervibe --skill rule-application
Please help me install this Agent Skill. Skill: rule-application Source: https://github.com/vTRKA/supervibe/tree/main/skills/rule-application Command: npx skills add https://github.com/vTRKA/supervibe --skill rule-application
It prevents AI-driven rule changes from being based on missing evidence or overly broad matches by requiring concrete source artifacts, validator/receipt proof, and scoped target verification during rule application.
Ask your AI tool to run rule-application in review mode for your codebase and return a rule-application-report with blocked reasons if any required evidence or receipts are missing.
High-intent search queries and answers about installing and using this skill.
Rule validation checks rule frontmatter, registry entries, and dry-run hit behavior against representative file matches before accepting applicability. This verifies rule loading and matching while isolating noisy rule text to prevent false positives.
A dry-run for rule application tests proposed rules against existing source artifacts to prove matching behavior without executing edits. You need it when reviewing new rules, changing existing audits, or adapting host adapters.
Protected-block safety checks verify scoped target spans for proposed edits by reviewing protected-span simplification risks. This isolates the smallest noisy rule text and avoids unsafe edits to protected code blocks.
Yes, rule-application runs in review mode to evaluate project rules against a codebase and returns a rule-application-report. This emits evidence, confidence levels, and a gate-on-exit status without modifying source files.
A rule-application-report contains evidence, confidence levels, and a gate-on-exit status with PASS, FAIL, PARTIAL, or BLOCKED results. It includes clear nextAction instructions instead of guessing applicability.
Overly broad matches occur when rule application lacks concrete source artifacts and validator proof. Requiring scoped rule construction, source-of-truth reads, and representative file matching via search prevents overreach.
majesticlabs-devGenerate hierarchical AGENTS.md structures with lightweight root files and sub-folder guidance.
jr2804Execute verification commands and analyze outputs before asserting completion.
AvneeshSarwateHost interactive UI components in Deno notebooks via WebSocket-bridged iframes.
blackplume233Automate rule-based cleanup of old or large files with dry-run previews.
