Skip to content

Latest commit

 

History

History
47 lines (37 loc) · 3.34 KB

File metadata and controls

47 lines (37 loc) · 3.34 KB

beagle-core

Language-agnostic developer workflows shared across every beagle plugin: commit and push, open pull requests, generate release notes, review implementation plans and repo structure, detect and fix LLM coding artifacts, process PR and reviewer feedback, and build new skills. Part of the beagle Agent Skills marketplace — see the full skill catalog.

Installation

For any coding agent that supports Agent Skills:

npx skills add existential-birds/beagle

For Claude Code:

claude plugin marketplace add https://github.com/existential-birds/beagle
claude plugin install beagle-core@existential-birds

Skills

Skill Description
commit-push Commit and push all local changes using Conventional Commits format
create-pr Create a pull request with a standardized description template
gen-release-notes Generate release notes for changes since a given tag
review-plan Review implementation plans for parallelization, TDD, types, libraries, and security before execution
review-structure Repo-wide structural-maintainability review — code-judo restructurings, 1k-line file guard, anti-spaghetti branching, canonical-layer enforcement, anti-magic abstractions, explicit type and boundary contracts
review-skill Review PRs that add or modify Agent Skills for structural validity, design quality, and marketplace consistency
skill-builder Create Agent Skills with best-practice structure, references, validation, and testing
subagent-prompt Hand off the current session's work to a fresh session as a portable orchestrator-plus-subagents prompt with per-task verification
prompt-improver Optimize prompts for code-related tasks following prompt-engineering best practices
receive-feedback Process external code review feedback with verification-first discipline, verifying claims before implementing and tracking disposition
fetch-pr-feedback Fetch unresolved review comments from a PR and evaluate them with the receive-feedback skill
respond-pr-feedback Post replies to PR review comments after evaluation and fixes
review-llm-artifacts Scan for LLM coding-agent artifacts across tests, dead code, abstraction, and style — changed files by default, --all for full-project
verify-llm-artifacts Adjudicate review-llm-artifacts findings as confirmed, false positive, or inconclusive before deletes
fix-llm-artifacts Apply fixes from a prior review-llm-artifacts run with safe/risky classification, respecting verification output
llm-artifacts-detection Reference: detection criteria for test quality, dead code, over-abstraction, and verbose LLM style
review-verification-protocol Reference: mandatory verification steps loaded before reporting any code review findings
review-feedback-schema Reference: schema for tracking code review outcomes to enable feedback-driven skill improvement
review-skill-improver Reference: analyzes feedback logs to surface patterns and suggest improvements to review skills

See Also