Agent Skills by pear-studio
Showing 21 vetted skills indexed across 1 GitHub repositories.
git-commit-brief
Standardize local git commit messages to Conventional Commits format.
backlog-implement
Guide implementer agents through verification, approval, and backlog closure.
review1-raise
Generate structured adversarial review findings from a local Git diff.
review3-confirm
Confirm Review-Reply items into consensus, deferment, denial, or clarification states.
review4-execute
Executes finalized consensus implementation entries from a review document after validation and tests.
backlog-check
Validate and prune obsolete entries in docs/dev/backlog.md.
backlog-add
Convert unstructured reports and notes into validated backlog entries.
dicepp-shell
Test DicePP chat commands in named sessions with fixed dice sequences.
onboard
Map DicePP repository structure and key file entry points for contributors.
review5-accept
Re-examine code diffs against a finalized review document for stage-5 acceptance.
toolchain-review
Reviews session context and writes a timestamped P0/P1/P2 severity report to .temp/report.
pr-create
Push a feature branch and generate a GitHub Pull Request from commit history.
test-quality-governor
Audit test suites and apply risk-aware refactors to protect behavior contracts.
review0-walkthrough
Walk through git diffs as logical units and record agreed decisions.
persona-inspect
Diagnose persona behavior by inspecting SQLite data and LLM trace records.
auto-test-guard
Enforce behavior-contract-focused assertions in software test suites.
start-feature
Create an isolated feature branch from master using git worktree.
run-tests
Run automated pytest test suites for the DicePP codebase.
review2-reply
Generate in-place Reply sections for each Rn in a review document.
pr-review
Review GitHub Pull Request diffs and generate structured code quality reports.
bump-version
Automate DicePP version increments with git commit and annotated tag preparation.